tomasz
/
code-examples
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Some my early projects from university times... Sentimental reasons

master
Tomasz Polgrabia 2025-03-09 11:55:28 +01:00
parent e9f413ef67
commit 47beddeb19
22 changed files with 1842 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
2014/2014_08/ada_server/.gitignore vendored Normal file
View File

@ -0,0 +1,3 @@
*.ali
*.o
server

7
2014/2014_08/ada_server/constants.ads Normal file
View File

@ -0,0 +1,7 @@
package constants is
cr : Character := Character'Val(13);
lf : Character := Character'Val(10);
newline: String := (cr, lf);
end constants;

75
2014/2014_08/ada_server/dispatchers.adb Normal file
View File

@ -0,0 +1,75 @@
package body dispatchers is
task body dispatcher is
HANDLERS_COUNT : constant Integer := 2;
MAX_SOCKETS_COUNT : constant Integer := 1000;
UNDISPATCHED_SOCKET_RETRY_TIME : constant Duration := 0.5;
handlers : array (1 .. HANDLERS_COUNT) of Handler;
type Socket_Index is range 1..MAX_SOCKETS_COUNT;
package Socket_Vector is new Ada.Containers.Vectors (
Element_Type => Socket_Type,
Index_Type => Socket_Index);
function Find_Free_Handler(s: Socket_Type) return Boolean is
begin
for id in handlers'Range loop
select
handlers(id).handle(s);
return True;
else
null;
end select;
end loop;
return False;
end Find_Free_Handler;
result : Boolean;
undispatched : Socket_Vector.Vector;
begin
accept start;
for id in handlers'Range loop
handlers(id).start(id);
end loop;
loop
select
accept dispatch (s: Socket_Type) do
Put_Line ("dispatch command");
result := Find_Free_Handler(s);
if not result then
Put_Line("All handlers are busy");
undispatched.Append(s);
end if;
end dispatch;
or
accept stop do
Put_Line("stop command, stopping dispatcher");
end stop;
exit;
or
delay UNDISPATCHED_SOCKET_RETRY_TIME;
if Integer(undispatched.Length) > 0 then
declare
s: Socket_Type := undispatched.Element(1);
begin
result := Find_Free_Handler(s);
if result then
undispatched.Delete(1);
Put_Line("Handled undispatched client");
end if;
end;
end if;
end select;
end loop;
for id in handlers'Range loop
handlers(id).stop;
end loop;
end dispatcher;
end dispatchers;

17
2014/2014_08/ada_server/dispatchers.ads Normal file
View File

@ -0,0 +1,17 @@
with Ada.Strings.Unbounded; use Ada.Strings.Unbounded;
with Ada.Text_IO; use Ada.Text_IO;
with Gnat.Sockets; use Gnat.Sockets;
with Ada.Calendar; use Ada.Calendar;
with Ada.Containers.Vectors;
with Handlers; use Handlers;
package dispatchers is
task dispatcher is
entry start;
-- signal to start
entry dispatch(s: Socket_Type);
-- we got accepted socket from the listener
entry stop;
-- signal to stop
end dispatcher;
end dispatchers;

247
2014/2014_08/ada_server/handlers.adb Normal file
View File

@ -0,0 +1,247 @@
package body handlers is
task body handler is
socket : Socket_Type;
sel : access Gnat.Sockets.Selector_Type := new
Gnat.Sockets.Selector_Type;
working : Boolean := True;
subtype Socket_Index is Integer range 1 .. 1000;
root_path : constant String := "/var/www";
package Socket_Vector is new Ada.Containers.Vectors (
Element_Type => Socket_Type,
Index_Type => Socket_Index
);
v : Socket_Vector.Vector;
task watcher is
entry start;
end watcher;
task body watcher is
R : Socket_Set_Type;
W : Socket_Set_Type;
E : Socket_Set_Type;
status : Selector_Status;
active : Socket_Type;
unparsed: Unbounded_String;
begin
accept start;
Put_Line("start watcher");
Create_Selector (sel.all);
loop
Put_Line("Watch loop start");
Empty(R);
Empty(W);
Empty(E);
Put("There is ");
Put(v.Last_Index, Width => 0);
Put_Line(" sockets to watch");
Put_Line("setting to read sockets");
for id in 1..v.Last_Index loop
set(R, v.Element(id));
end loop;
Put_Line("start of check_selector");
Check_Selector(
sel.all,
R,
W,
E,
status,
Gnat.Sockets.Forever);
Put_Line("end of check_selector");
-- wychodzimy jedynie wtedy, gdy zmienna working
-- jest False. Inaczej to jedynie przeładowanie
-- socketów np. dodano nam socket do obserwacji
exit when status = Gnat.Sockets.Aborted
and not working;
if status /= Gnat.Sockets.Aborted then
Put_Line("Sending command to client handler");
Get(R, active);
Handler.command(active);
end if;
Put_Line("Watch loop end");
end loop;
Close_Selector (sel.all);
Put_Line("exit watcher");
end watcher;
procedure Handle_File(s: Socket_Type;
path: String;
finished: out Boolean) is
sz : Natural := Natural(Size(path));
subtype Content_Type is String (1 .. sz);
package dio is new Ada.Direct_IO(Content_Type);
File : dio.File_Type;
contents : Content_Type;
channel : Stream_Access := Stream(s);
begin
Put_Line("Handle_File start");
Put_Line("Path: " & path);
Put_Line("File size: " & Natural'Image(sz));
dio.Open(
File => File,
Mode => dio.In_File,
Name => path);
Put_Line("File opened");
String'Write(channel, "HTTP/1.0 200 OK" & newline);
String'Write(channel, "Server: Ada Server 0.1" &
newline);
Put_Line("Root path: " & root_path & path);
Put_Line("File size: " & Natural'Image(sz));
dio.Read (File, Item => contents);
Put_Line("Contents: " & contents & newline);
String'Write(channel, newline);
String'Write(channel, contents);
dio.Close(File);
Put_Line("Closed file");
Close_Socket(s);
Put_Line("Closed socket");
finished := True;
end Handle_File;
procedure Handle_Socket(s: Socket_Type; finished: out Boolean) is
subtype Line_Type is String (1 .. 4096);
channel : Stream_Access := Stream (s);
line : Line_Type;
elem_line : Stream_Element_Array (1 .. 4096);
last : Stream_Element_Offset;
package af renames Ada.Strings.Fixed;
function Convert is new Ada.Unchecked_Conversion (
Source => Stream_Element_Array,
Target => Line_Type
);
begin
Put_Line("Got command to execute");
channel := Stream (s);
Receive_Socket(s, elem_line, last);
line := Convert(elem_line);
Put("Odczytano: ");
Put(Integer(last), Width => 0);
Put_Line(" znakow");
if last >= 4 and line (1 .. 4) = "GET " then
Put_Line("GET Command: " & line);
declare
EOL : Integer := af.Index(line, newline);
HTTP_Version_Offset : Integer := af.Index(line, "HTTP/1.");
path : String := line (5..HTTP_Version_Offset - 2);
begin
if Exists (root_path & path) then
Put_Line("Kind: " & File_Kind'Image(Kind(root_path & path)));
if Kind(root_path & path) = Ordinary_File then
Put_Line("Full path: '" & root_path & path & "'");
handle_file(s,
root_path & path,
finished);
return;
else
-- pewnie katalog
if Exists(root_path & path & "index.html") then
handle_file(s,
root_path & path &
"index.html",
finished);
return;
end if;
if Exists(root_path & path & "index.htm") then
handle_file(s,
root_path & path &
"index.htm",
finished);
return;
end if;
String'Write(channel, "HTTP/1.0 200 OK" & newline);
String'Write(channel, "Server: Ada Server 0.1" &
newline & newline);
String'Write(channel, "<html><body>" &
"directory</body></html>" & newline);
end if;
else
String'Write (channel, "HTTP/1.0 404 NOT FOUND" &
newline);
String'Write (channel, "Server: Ada Server 0.1" &
newline & newline);
String'Write (channel, "<html><body>" &
"404 Error</body></html>");
end if;
end;
else
String'Write(channel, "Unknown" & newline);
end if;
Put_Line("Closing socket");
Close_Socket(s);
finished := True;
end Handle_Socket;
idx : Integer;
begin
accept start(id: Integer) do
idx := id;
end start;
Watcher.Start;
Put("start client handler with id: ");
Put(idx, Width => 0);
New_Line;
loop
select
accept handle(s: Socket_Type) do
socket := s;
v.Append(s);
end handle;
Abort_Selector(sel.all);
Put_Line("Socket consumed by client handler");
or
accept command(s: Socket_Type) do
declare
finished: Boolean;
begin
Handle_Socket(s, finished);
if finished then
v.Delete(v.Find_Index(s));
end if;
end;
end command;
or
accept stop do
Put_Line("Got command to stop");
working := False;
Abort_Selector(sel.all);
end stop;
exit;
or
terminate;
end select;
end loop;
Put_Line("end client handler");
end handler;
end handlers;

37
2014/2014_08/ada_server/handlers.ads Normal file
View File

@ -0,0 +1,37 @@
with Ada.Containers.Vectors;
with Ada.Integer_Text_IO; use Ada.Integer_Text_IO;
with Ada.Strings.Unbounded; use Ada.Strings.Unbounded;
with Ada.Text_IO; use Ada.Text_IO;
with GNAT.Sockets; use GNAT.Sockets;
with Constants; use Constants;
with Ada.Calendar; use Ada.Calendar;
with Ada.Streams; use Ada.Streams;
with Ada.Unchecked_Conversion;
with Ada.Strings.Fixed;
with Ada.Directories; use Ada.Directories;
with Ada.Direct_IO;
with Ada.Exceptions; use Ada.Exceptions;
with Ada.IO_Exceptions; use Ada.IO_Exceptions;
package handlers is
type Client_Type is record
socket : Socket_Type;
read : Unbounded_String;
unread : Unbounded_String;
last_active : Time;
end record;
type Client_Index_Type is range 1..100;
package Client_Vector_Type is new Ada.Containers.Vectors (
Element_Type => Client_Type,
Index_Type => Client_Index_Type);
task type handler is
entry start(id: Integer);
entry handle(s: Socket_Type);
entry command(s: Socket_Type);
entry stop;
end handler;
end handlers;

67
2014/2014_08/ada_server/server.adb Normal file
View File

@ -0,0 +1,67 @@
with Ada.Text_IO; use Ada.Text_IO;
with Ada.Integer_Text_IO; use Ada.Integer_Text_IO;
with GNAT.Sockets; use GNAT.Sockets;
with signals; use signals;
with Ada.Containers.Vectors;
with Handlers; use Handlers;
with Constants; use Constants;
with Dispatchers; use Dispatchers;
procedure server is
HOST : constant String := "localhost";
PORT : constant Port_Type := 6666;
work : Boolean := True;
address : Sock_Addr_Type;
server : Socket_Type;
socket : Socket_Type;
h2 : Sigint_Handler;
status : Gnat.Sockets.Selector_Status;
selector: signals.selector_Access;
req : Request_Type(Non_Blocking_IO);
begin
Put_Line("start of server");
Dispatcher.Start;
selector := new Gnat.Sockets.Selector_Type;
h2.selector(selector);
Gnat.Sockets.Create_Selector(selector.all);
address.Addr := Addresses (Get_Host_By_Name(host), 1);
address.Port := PORT;
Create_Socket(server);
Set_Socket_Option (
server,
Socket_Level,
(Reuse_Address, True));
Bind_Socket (server, address);
Put_Line("Binded");
Listen_Socket(server);
Put_Line("Listening");
loop
Put_Line("Waiting to accept");
Accept_Socket (
Server => server,
Socket => socket,
Address => address,
Timeout => GNAT.Sockets.Forever,
Selector => selector,
Status => status);
exit when status = Gnat.Sockets.Aborted;
Put_Line("Accepted");
Control_Socket(socket, req);
Dispatcher.dispatch(socket);
Put_Line("Handled by listener");
end loop;
Gnat.Sockets.Close_Selector(selector.all);
Close_Socket (server);
Dispatcher.Stop;
Put_Line("end of server");
end server;

16
2014/2014_08/ada_server/signals.adb Normal file
View File

@ -0,0 +1,16 @@
package body signals is
protected body Sigint_Handler is
procedure selector(s: Selector_Access) is
begin
sel := s;
end;
procedure Handle is
begin
Call_Count := Call_Count + 1;
Put_Line("SIGINT handled");
Gnat.Sockets.Abort_Selector(sel.all);
end Handle;
end Sigint_Handler;
end signals;

23
2014/2014_08/ada_server/signals.ads Normal file
View File

@ -0,0 +1,23 @@
with Ada.Interrupts; use Ada.Interrupts;
with Ada.Interrupts.Names; use Ada.Interrupts.Names;
with Ada.Text_IO; use Ada.Text_IO;
with Gnat.Sockets; use Gnat.Sockets;
package signals is
pragma Unreserve_All_Interrupts;
type Selector_Access is access all Gnat.Sockets.Selector_Type;
protected type Sigint_Handler is
procedure selector(s: Selector_Access);
procedure Handle;
pragma Interrupt_Handler(Handle);
pragma Attach_Handler(Handle, Sigint);
private
Call_Count : Natural := 0;
sel : Selector_Access;
end Sigint_Handler;
end signals;

1
2015/2015_03/shoe-soles/README.md Normal file
View File

@ -0,0 +1 @@
Projekt rozpoznawania obuwia. Bardzo wczesna faza.

75
2015/2015_03/shoe-soles/dect.py Normal file
View File

@ -0,0 +1,75 @@
from scipy import ndimage
from skimage import filters,feature
import sys
import matplotlib.pyplot as plt
import math
import numpy as np
import pdb
def calculate_edges(img):
# TODO fix the performance - added ad hoc
(n,m) = img.shape
edge = np.zeros((n,m), dtype=np.bool)
# pdb.set_trace()
for i in range(0, n):
for j in range(0, m):
edge_point = False
for k in range(-1, 2):
if edge_point:
break
for l in range(-1,2):
if edge_point:
break
y = (i+k)%n
x = (j+l)%m
if img[y,x] != img[i,j]:
edge_point = True
break
edge[i,j] = edge_point
return edge
def calc_shape_corners(img_mask2, range_ext=8):
(my,mx) = ndimage.center_of_mass(img_mask2)
# edge = feature.canny(img_mask2)
edge = calculate_edges(img_mask2)
(n,m) = img_mask2.shape
print("Center of mass {}x{}".format(mx,my))
distances = np.zeros((n,m), dtype=np.float)
distances[:,:] = 0
for i in range(0, n):
for j in range(0, m):
if edge[i,j]:
dx = j - mx
dy = i - my
d = math.sqrt(dx*dx+dy*dy)
distances[i,j] = d
max_distances = ndimage.maximum_filter(distances, range_ext)
diff = np.abs(max_distances - distances)
corners = edge * (diff < 1e-3)
return corners
if __name__ == "__main__":
img = ndimage.imread(sys.argv[1])
img_gray = 0.21 * img[:,:,0] + 0.72 * img[:,:,1] + 0.07 * img[:,:,2]
otsu_lvl = filters.threshold_otsu(img_gray)
img_mask = img_gray <= otsu_lvl
img_mask2 = ndimage.median_filter(img_mask, 3)
(n,m) = img_mask2.shape
corners = calc_shape_corners(img_mask2)
mat_x = np.tile(np.arange(0,m),n).reshape((n,m))
mat_y = np.tile(np.arange(0,n),m).reshape((m,n)).T
xs = mat_x[corners]
ys = mat_y[corners]
plt.imshow(img_gray, cmap="Greys_r")
plt.plot(xs, ys, "r+")
plt.show()

8
2015/2015_03/shoe-soles/filtering-print.py Normal file
View File

@ -0,0 +1,8 @@
from scipy import ndimage
import matplotlib.pyplot as plt
img = ndimage.imread("/home/tpolgrabia/Pobrane/CSFID/tracks_cropped/00003.jpg")
img_filtered = ndimage.median_filter(img, 5)
plt.imshow(img_filtered, cmap="Greys_r")
plt.show()

50
2015/2015_03/shoe-soles/image_boundaries.py Normal file
View File

@ -0,0 +1,50 @@
# -*- coding: UTF-8 -*-
import numpy as np
def find_minx(mat,val = 1):
mat2 = mat.reshape((-1,1), order='F')
(n,m) = mat.shape
n2 = mat2.shape[0]
for (el,idx) in zip(np.nditer(mat2),range(0,n2)):
if el == val:
return idx / n
return None
def find_miny(mat,val = 1):
mat2 = mat.reshape((1,-1))
(n,m) = mat.shape
n2 = mat2.shape[1]
for (el,idx) in zip(np.nditer(mat2),range(0,n2)):
if el == val:
return idx / m
return None
def find_maxx(mat,val = 1):
mat2 = np.rot90(np.rot90(mat)).reshape((-1,1), order='F')
(n,m) = mat.shape
n2 = mat2.shape[0]
for (el,idx) in zip(np.nditer(mat2),range(0,n2)):
if el == val:
return m - 1 - idx / n
return None
def find_maxy(mat,val = 1):
mat_rotated = np.rot90(np.rot90(mat))
mat2 = mat_rotated.reshape((1,-1))
(n,m) = mat.shape
n2 = mat2.shape[1]
for (el,idx) in zip(np.nditer(mat2),range(0,n2)):
if el == val:
return n - 1 - idx / m
return None
def calc_image_crop(img, val = 255):
minx = find_minx(img,val)
maxx = find_maxx(img,val)
miny = find_miny(img,val)
maxy = find_maxy(img,val)
return (minx,maxx,miny,maxy)

66
2015/2015_03/shoe-soles/image_boundaries2.py Normal file
View File

@ -0,0 +1,66 @@
# -*- coding: UTF-8 -*-
import numpy as np
from scipy import ndimage
def find_miny(mat,val = 1):
(n,m) = mat.shape
for i in range(0,n):
for j in range(0,m):
if mat[i,j] == val:
return i
return None
def find_minx(mat,val = 1):
(n,m) = mat.shape
for j in range(0,m):
for i in range(0,n):
if mat[i,j] == val:
return j
return None
def find_maxy(mat,val = 1):
(n,m) = mat.shape
for i in range(n-1,-1,-1):
for j in range(m-1,-1,-1):
if mat[i,j] == val:
return i
return None
def find_maxx(mat,val = 1):
(n,m) = mat.shape
for j in range(m-1,-1,-1):
for i in range(n-1,-1,-1):
if mat[i,j] == val:
return j
return None
def calc_image_crop(img, val = 255):
minx = find_minx(img,val)
maxx = find_maxx(img,val)
miny = find_miny(img,val)
maxy = find_maxy(img,val)
return (minx,maxx,miny,maxy)
def crop_image(img,val = 255):
(minx,maxx,miny,maxy) = calc_image_crop(img,val)
return img[miny:maxy,minx:maxx]
def crop_image_prefilled(img,val = 255):
(minx,maxx,miny,maxy) = calc_image_crop(img,val)
img_filled = ndimage.binary_fill_holes(img)
return img_filled[miny:maxy,minx:maxx]
def dimension_rate(img):
(n,m) = img.shape
return float(m)/float(n)
def image_to_mask(img,threshold = 128):
m = img >= threshold
return 1*m

85
2015/2015_03/shoe-soles/measures.py Normal file
View File

@ -0,0 +1,85 @@
#!/usr/bin/python
from scipy import ndimage
import sys
import math
import matplotlib.pyplot as plt
import numpy as np
print("Name: {}".format(__name__))
if __name__ != "__main__":
print("not runned")
sys.exit(0)
if len(sys.argv) <= 1:
print("Too small arguments")
sys.exit(1)
path = sys.argv[1]
img = ndimage.imread(path)
img_gray = img[:,:,0]
(n,m) = img_gray.shape
mask = img_gray >= 128
sx = 0.0
sy = 0.0
c = 0
for i in range(0, n):
for j in range(0, m):
if mask[i,j]:
sx += j
sy += i
c += 1
ax = sx / c
ay = sy / c
print("Position: {} x {}".format(ax,ay))
grav_measure = 0.0
for i in range(0, n):
for j in range(0, m):
if mask[i,j]:
grav_measure += (j - ax) * (j - ax) + (i - ay) * (i - ay)
grav_measure2 = math.sqrt(grav_measure / c)
print("Grav measure: {}".format(grav_measure))
# vector of offset
ox = 0.0
oy = 0.0
for i in range(0, n):
for j in range(0, m):
if mask[i,j]:
ox += math.pow(j - ax, 1.0)
oy += math.pow(i - ay, 1.0)
print("Vector of offset (measure of beaing simetrical): {}x{}".format(ox,oy))
img_filled = ndimage.binary_fill_holes(img_gray)
s = np.sum(img_filled)
unit = max(n,m)
c2 = c / float(unit)
s2 = s / (float(unit)*float(unit))
print("l: {}, s: {}".format(c2,s2))
# plt.imshow(img_filled, cmap="Greys_r")
# plt.show()
circularity1 = 2.0 * math.sqrt(s2 / np.pi)
print("Circularity1: {}".format(circularity1))
circularity2 = c2 / math.pi
print("Circularity2: {}".format(circularity2))
w_measure = c2 / (2.0 * math.sqrt(np.pi * s2)) - 1.0
print("W measure: {}".format(w_measure))
w9_measure = 2.0 * math.sqrt(np.pi * s2) / c2
print("W9 measure (malinkowskiej) {}".format(w9_measure))

437
2015/2015_03/shoe-soles/old/ipython_log.py Normal file
View File

@ -0,0 +1,437 @@
# IPython log file
from dect import calc_shape_corners
calc_shape_corners(img)
from scipy import ndimage
img = ndimage.imread("/home/tpolgrabia/Pobrane/CSFID/tracks_cropped/00003.jpg")
img.shape
#[Out]# (570, 193)
import matplotlib.pyplot as plt
plt.imshow(img, cmap="Greys_r")
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4e9664d0>
plt.show()
from skimage import filters
otsu_lvl = filters.threshold_otsu(img)
otsu_lvl
#[Out]# 139
plt.imshow(img, cmap="Greys_r")
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4dc30790>
plt.show()
plt.imshow(img >= otsu_lvl, cmap="Greys_r")
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4db63690>
plt.show()
m = img <= otsu_lvl
plt.imshow(m, cmap="Greys_r")
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4da97250>
plt.show()
img_label, nb_labels = ndimage.label(m)
nb_labels
#[Out]# 571
plt.imshow(nb_labels)
plt.imshow(img_label))
plt.imshow(img_label)
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4e921250>
plt.show()
sizes = ndimage.sum(m, img_label, range(0, nb_labels+1))
sizes
#[Out]# array([ 0.00000000e+00, 1.00000000e+00, 4.50000000e+01,
#[Out]# 1.70000000e+01, 6.20000000e+01, 1.50000000e+01,
#[Out]# 5.00000000e+00, 2.00000000e+00, 2.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 3.00000000e+00, 9.00000000e+00,
#[Out]# 2.90000000e+01, 1.00000000e+00, 5.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 4.00000000e+00, 5.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 2.40000000e+01, 6.00000000e+00,
#[Out]# 5.00000000e+00, 3.10000000e+01, 1.00000000e+00,
#[Out]# 5.30000000e+01, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 2.00000000e+00, 2.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 6.00000000e+00,
#[Out]# 3.80000000e+01, 1.00000000e+00, 3.00000000e+00,
#[Out]# 3.00000000e+00, 1.20000000e+01, 1.10000000e+01,
#[Out]# 1.00000000e+01, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 5.60000000e+01, 3.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 2.50000000e+01,
#[Out]# 1.00000000e+00, 3.00000000e+00, 4.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 3.00000000e+00, 3.00000000e+00,
#[Out]# 3.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 3.00000000e+00, 2.20000000e+01,
#[Out]# 1.00000000e+00, 3.00000000e+00, 3.00000000e+00,
#[Out]# 1.00000000e+00, 8.00000000e+00, 1.00000000e+01,
#[Out]# 1.00000000e+00, 2.00000000e+00, 1.46000000e+02,
#[Out]# 2.00000000e+00, 3.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.40000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 4.00000000e+00, 1.15550000e+04,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 5.80000000e+01, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 8.00000000e+00,
#[Out]# 1.00000000e+01, 2.00000000e+00, 3.00000000e+00,
#[Out]# 1.00000000e+00, 3.00000000e+00, 5.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 5.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 3.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 5.40000000e+01, 5.00000000e+00,
#[Out]# 8.00000000e+00, 1.90000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+01, 3.00000000e+00, 2.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 3.30000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 4.00000000e+00, 1.00000000e+00,
#[Out]# 4.00000000e+00, 1.30000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 2.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 3.00000000e+00,
#[Out]# 1.40000000e+01, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 3.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 5.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 5.00000000e+00, 4.00000000e+00,
#[Out]# 4.00000000e+00, 2.20000000e+01, 2.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 3.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 4.00000000e+00,
#[Out]# 4.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 5.00000000e+00, 1.10000000e+01,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 3.00000000e+00, 4.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 4.00000000e+00, 5.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.20000000e+01, 1.01400000e+03,
#[Out]# 4.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 5.00000000e+00,
#[Out]# 4.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.90000000e+01, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+01, 3.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 4.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.10000000e+01, 5.00000000e+00, 1.00000000e+00,
#[Out]# 7.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 6.40000000e+01, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 4.00000000e+00, 5.80000000e+01, 9.00000000e+00,
#[Out]# 9.30000000e+01, 8.10000000e+01, 6.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 2.91000000e+02,
#[Out]# 3.00000000e+00, 1.00000000e+00, 4.00000000e+00,
#[Out]# 1.30000000e+01, 1.35000000e+02, 3.10000000e+01,
#[Out]# 4.00000000e+00, 2.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 3.00000000e+01, 1.00000000e+01,
#[Out]# 5.20000000e+01, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 4.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 2.62000000e+02, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 4.00000000e+00, 2.28000000e+03, 1.00000000e+00,
#[Out]# 1.39000000e+02, 2.00000000e+00, 2.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 2.75000000e+02,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 8.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.13000000e+02, 2.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 2.27000000e+02,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.40000000e+01, 2.00000000e+00,
#[Out]# 2.26000000e+02, 2.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.54000000e+02, 1.00000000e+00, 1.10000000e+01,
#[Out]# 1.00000000e+00, 1.00000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 9.00000000e+00, 1.00000000e+00,
#[Out]# 1.40000000e+01, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.11000000e+02, 2.00000000e+00, 1.00000000e+00,
#[Out]# 5.00000000e+00, 6.00000000e+00, 1.89000000e+02,
#[Out]# 1.00000000e+00, 8.00000000e+00, 1.00000000e+00,
#[Out]# 7.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 7.90000000e+01, 3.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 5.00000000e+00, 1.00000000e+00,
#[Out]# 8.00000000e+00, 2.00000000e+00, 1.94000000e+02,
#[Out]# 1.20000000e+01, 1.00000000e+00, 4.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 8.70000000e+01, 1.55000000e+02, 1.68400000e+03,
#[Out]# 3.00000000e+00, 1.30000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 6.00000000e+00, 1.87900000e+03, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 6.00000000e+00,
#[Out]# 3.15000000e+02, 1.59000000e+02, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 5.00000000e+00,
#[Out]# 6.80000000e+01, 2.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.59000000e+02, 3.95000000e+02, 6.00000000e+00,
#[Out]# 5.00000000e+00, 6.00000000e+00, 6.80000000e+01,
#[Out]# 4.00000000e+01, 5.10000000e+01, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 3.69000000e+02, 1.17000000e+02, 1.00000000e+00,
#[Out]# 2.00000000e+00, 3.80000000e+01, 5.61000000e+02,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 9.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 1.00000000e+00,
#[Out]# 2.89000000e+02, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 2.00000000e+00,
#[Out]# 1.70000000e+01, 2.00000000e+00, 1.00000000e+00,
#[Out]# 4.00000000e+00, 3.00000000e+00, 4.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 4.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00, 3.00000000e+00,
#[Out]# 3.00000000e+00, 2.00000000e+00, 6.00000000e+00,
#[Out]# 1.00000000e+00, 3.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 2.00000000e+00, 3.10000000e+01,
#[Out]# 7.00000000e+00, 2.00000000e+00, 9.00000000e+00,
#[Out]# 6.10000000e+01, 1.00000000e+00, 6.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 7.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 6.00000000e+00,
#[Out]# 2.40000000e+01, 5.00000000e+00, 2.00000000e+00,
#[Out]# 1.00000000e+00, 2.00000000e+00, 9.00000000e+00,
#[Out]# 9.40000000e+01, 1.00000000e+00, 1.00000000e+00,
#[Out]# 2.00000000e+00, 1.00000000e+00, 8.00000000e+00,
#[Out]# 2.60000000e+01, 1.00000000e+00, 2.50000000e+01,
#[Out]# 1.00000000e+00, 1.00000000e+00, 4.00000000e+00,
#[Out]# 1.80000000e+01, 2.00000000e+00, 7.00000000e+00,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.40000000e+01, 1.40000000e+01, 4.90000000e+01,
#[Out]# 1.00000000e+00, 2.00000000e+00, 4.00000000e+00,
#[Out]# 1.00000000e+00, 2.50000000e+01, 4.00000000e+00,
#[Out]# 4.00000000e+00, 7.00000000e+00, 2.30000000e+01,
#[Out]# 3.00000000e+00, 5.00000000e+00, 8.00000000e+00,
#[Out]# 6.60000000e+01, 8.00000000e+00, 1.00000000e+01,
#[Out]# 3.00000000e+00, 6.00000000e+00, 2.00000000e+00,
#[Out]# 4.00000000e+00, 1.60000000e+01, 1.00000000e+00,
#[Out]# 6.70000000e+01, 7.00000000e+00, 1.50000000e+02,
#[Out]# 4.10000000e+01, 1.00000000e+00, 2.20000000e+01,
#[Out]# 3.00000000e+00, 1.00000000e+00, 1.00000000e+00,
#[Out]# 1.00000000e+00, 1.00000000e+00])
size_min_limit = 50
remove_segments = sizes < size_min_limit
remove_segments[img_label]
#[Out]# array([[ True, True, True, ..., True, True, True],
#[Out]# [ True, True, True, ..., True, True, True],
#[Out]# [ True, True, True, ..., True, True, True],
#[Out]# ...,
#[Out]# [False, False, False, ..., True, True, True],
#[Out]# [False, False, False, ..., True, True, True],
#[Out]# [False, False, False, ..., True, True, True]], dtype=bool)
remove_mask = remove_segments[img_label]
img_label[remove_mask] = 0
plt.imshow(img_label)
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a4d804d50>
plt.show()
import numpy as np
get_ipython().magic(u'pinfo np.unique')
labels = np.unique(img_label)
labels
#[Out]# array([ 0, 4, 30, 54, 83, 92, 97, 127, 239, 276, 289, 291, 292,
#[Out]# 296, 301, 309, 319, 325, 327, 338, 351, 356, 363, 369, 381, 386,
#[Out]# 393, 401, 408, 409, 410, 430, 435, 436, 441, 447, 448, 452, 454,
#[Out]# 459, 460, 464, 474, 504, 519, 552, 561, 563], dtype=int32)
get_ipython().magic(u'pinfo np.searchsorted')
img_label = np.searchsorted(labels, img_label)
nb_labels = len(labels)
plt.imshow(img_label)
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a85719790>
plt.show()
ndimage.center_of_mass(img_label, labels, nb_labels+1)
ndimage.center_of_mass(img_label, labels, nb_labels+1)
get_ipython().magic(u'pinfo ndimage.center_of_mass')
ndimage.center_of_mass(m, img_label, labels)
#[Out]# [(274.99289099526067, 99.926540284360186), (60.863013698630134, 45.006849315068493), (413.25593824228031, 114.4667458432304), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan)]
centers = ndimage.center_of_mass(m, img_label, labels)
centers
#[Out]# [(274.99289099526067, 99.926540284360186), (60.863013698630134, 45.006849315068493), (413.25593824228031, 114.4667458432304), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan), (nan, nan)]
centers[0]
#[Out]# (274.99289099526067, 99.926540284360186)
centers[1]
#[Out]# (60.863013698630134, 45.006849315068493)
centers[2]
#[Out]# (413.25593824228031, 114.4667458432304)
centers[3]
#[Out]# (nan, nan)
centers[4]
#[Out]# (nan, nan)
centers[5]
#[Out]# (nan, nan)
labels]
labels
#[Out]# array([ 0, 4, 30, 54, 83, 92, 97, 127, 239, 276, 289, 291, 292,
#[Out]# 296, 301, 309, 319, 325, 327, 338, 351, 356, 363, 369, 381, 386,
#[Out]# 393, 401, 408, 409, 410, 430, 435, 436, 441, 447, 448, 452, 454,
#[Out]# 459, 460, 464, 474, 504, 519, 552, 561, 563], dtype=int32)
labels = np.unique(img_label)
labels
#[Out]# array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
#[Out]# 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,
#[Out]# 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47])
centers = ndimage.center_of_mass(m, img_label, labels)
centers
#[Out]# [(274.99289099526067, 99.926540284360186), (7.112903225806452, 30.677419354838708), (28.69811320754717, 103.75471698113208), (42.089285714285715, 112.85714285714286), (60.863013698630134, 45.006849315068493), (154.04898312418865, 112.49026395499784), (71.5, 32.413793103448278), (89.5, 60.388888888888886), (184.49112426035504, 34.280078895463511), (222.828125, 47.25), (238.68965517241378, 186.10344827586206), (241.40860215053763, 52.473118279569896), (241.8641975308642, 121.19753086419753), (250.85910652920961, 163.26116838487974), (259.62222222222221, 56.814814814814817), (259.84615384615387, 121.51923076923077), (269.1641221374046, 156.87022900763358), (445.33201754385965, 185.51271929824563), (278.58992805755395, 67.2158273381295), (286.66181818181821, 153.59272727272727), (297.10619469026551, 74.13274336283186), (305.29955947136563, 151.12775330396477), (318.63274336283183, 77.853982300884951), (327.37853107344631, 144.00564971751413), (342.48648648648651, 78.090090090090087), (350.68783068783068, 140.42328042328043), (360.32911392405066, 78.455696202531641), (368.29381443298968, 140.70103092783506), (377.85057471264366, 74.402298850574709), (386.11612903225807, 141.27741935483871), (413.25593824228031, 114.4667458432304), (495.27833954230977, 6.3432676955827567), (443.28888888888889, 145.54920634920634), (442.47169811320754, 99.345911949685529), (448.38235294117646, 61.676470588235297), (460.94339622641508, 100.12578616352201), (462.9594936708861, 144.1113924050633), (466.64705882352939, 60.441176470588232), (476.21568627450978, 175.94117647058823), (481.65853658536588, 143.49864498644988), (479.21367521367523, 99.282051282051285), (507.28163992869878, 96.748663101604279), (500.46020761245677, 144.54325259515571), (523.13114754098365, 145.36065573770492), (529.47872340425533, 126.72340425531915), (548.36363636363637, 83.015151515151516), (554.2388059701492, 81.880597014925371), (563.0333333333333, 103.40000000000001)]
centers[:,0]
np.array(centers)
#[Out]# array([[ 274.992891 , 99.92654028],
#[Out]# [ 7.11290323, 30.67741935],
#[Out]# [ 28.69811321, 103.75471698],
#[Out]# [ 42.08928571, 112.85714286],
#[Out]# [ 60.8630137 , 45.00684932],
#[Out]# [ 154.04898312, 112.49026395],
#[Out]# [ 71.5 , 32.4137931 ],
#[Out]# [ 89.5 , 60.38888889],
#[Out]# [ 184.49112426, 34.2800789 ],
#[Out]# [ 222.828125 , 47.25 ],
#[Out]# [ 238.68965517, 186.10344828],
#[Out]# [ 241.40860215, 52.47311828],
#[Out]# [ 241.86419753, 121.19753086],
#[Out]# [ 250.85910653, 163.26116838],
#[Out]# [ 259.62222222, 56.81481481],
#[Out]# [ 259.84615385, 121.51923077],
#[Out]# [ 269.16412214, 156.87022901],
#[Out]# [ 445.33201754, 185.5127193 ],
#[Out]# [ 278.58992806, 67.21582734],
#[Out]# [ 286.66181818, 153.59272727],
#[Out]# [ 297.10619469, 74.13274336],
#[Out]# [ 305.29955947, 151.1277533 ],
#[Out]# [ 318.63274336, 77.8539823 ],
#[Out]# [ 327.37853107, 144.00564972],
#[Out]# [ 342.48648649, 78.09009009],
#[Out]# [ 350.68783069, 140.42328042],
#[Out]# [ 360.32911392, 78.4556962 ],
#[Out]# [ 368.29381443, 140.70103093],
#[Out]# [ 377.85057471, 74.40229885],
#[Out]# [ 386.11612903, 141.27741935],
#[Out]# [ 413.25593824, 114.46674584],
#[Out]# [ 495.27833954, 6.3432677 ],
#[Out]# [ 443.28888889, 145.54920635],
#[Out]# [ 442.47169811, 99.34591195],
#[Out]# [ 448.38235294, 61.67647059],
#[Out]# [ 460.94339623, 100.12578616],
#[Out]# [ 462.95949367, 144.11139241],
#[Out]# [ 466.64705882, 60.44117647],
#[Out]# [ 476.21568627, 175.94117647],
#[Out]# [ 481.65853659, 143.49864499],
#[Out]# [ 479.21367521, 99.28205128],
#[Out]# [ 507.28163993, 96.7486631 ],
#[Out]# [ 500.46020761, 144.5432526 ],
#[Out]# [ 523.13114754, 145.36065574],
#[Out]# [ 529.4787234 , 126.72340426],
#[Out]# [ 548.36363636, 83.01515152],
#[Out]# [ 554.23880597, 81.88059701],
#[Out]# [ 563.03333333, 103.4 ]])
centers = np.array(centers)
centers[:,0]
#[Out]# array([ 274.992891 , 7.11290323, 28.69811321, 42.08928571,
#[Out]# 60.8630137 , 154.04898312, 71.5 , 89.5 ,
#[Out]# 184.49112426, 222.828125 , 238.68965517, 241.40860215,
#[Out]# 241.86419753, 250.85910653, 259.62222222, 259.84615385,
#[Out]# 269.16412214, 445.33201754, 278.58992806, 286.66181818,
#[Out]# 297.10619469, 305.29955947, 318.63274336, 327.37853107,
#[Out]# 342.48648649, 350.68783069, 360.32911392, 368.29381443,
#[Out]# 377.85057471, 386.11612903, 413.25593824, 495.27833954,
#[Out]# 443.28888889, 442.47169811, 448.38235294, 460.94339623,
#[Out]# 462.95949367, 466.64705882, 476.21568627, 481.65853659,
#[Out]# 479.21367521, 507.28163993, 500.46020761, 523.13114754,
#[Out]# 529.4787234 , 548.36363636, 554.23880597, 563.03333333])
xs = centers[:,0]
ys = centers[:,0]
xs = centers[:,1]
plt.imshow(m, cmap="Greys_r")
#[Out]# <matplotlib.image.AxesImage object at 0x7f7a84e8af10>
plt.plot(xs, ys, "r+")
#[Out]# [<matplotlib.lines.Line2D object at 0x7f7a84efad10>]
plt.show()
get_ipython().magic(u'logstart')
get_ipython().magic(u'pwd ')
#[Out]# u'/home/tpolgrabia/Dokumenty/prace/signal-processing/shoe-soles'
help %logstart
get_ipython().magic(u'pinfo %logstart')
get_ipython().magic(u'logstart -o')
get_ipython().magic(u'logstop')
get_ipython().magic(u'logstart -o')
get_ipython().magic(u'logstop')
# nie, 10 lip 2016 18:10:54
get_ipython().magic(u'pwd ')
#[Out]# u'/home/tpolgrabia/Dokumenty/prace/signal-processing/shoe-soles'
# nie, 10 lip 2016 18:10:59
get_ipython().system(u'cat ipython_log.py')
# nie, 10 lip 2016 18:11:52
pw
# nie, 10 lip 2016 18:12:19
get_ipython().magic(u'edit')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n corners += seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:16:31
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n corners += seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:17:03
get_ipython().magic(u'edit dect.py')
# nie, 10 lip 2016 18:17:28
get_ipython().magic(u'edit __')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n corners += seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:18:12
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:18:50
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n plt.imshow(seg_corner, cmap="Greys_r")\n plt.show()\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:20:23
get_ipython().magic(u'edit __')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n plt.imshow(mseg, cmap="Greys_r")\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:20:44
get_ipython().magic(u'edit __')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n plt.imshow(seg_corner, cmap="Greys_r")\n plt.show()\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:21:07
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n plt.imshow(mseg, cmap="Greys_r")\n plt.show()\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:22:01
get_ipython().magic(u'edit test.py')
# nie, 10 lip 2016 18:22:10
get_ipython().magic(u'save 1-92 try.py')
# nie, 10 lip 2016 18:22:16
get_ipython().magic(u'save try.py 1-92')
# nie, 10 lip 2016 18:22:22
get_ipython().magic(u'edit test.py')
# nie, 10 lip 2016 18:22:33
get_ipython().magic(u'edit ___')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n# plt.imshow(mseg, cmap="Greys_r")\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:22:58
get_ipython().magic(u'edit ___')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n seg_corner = calc_shape_corners(mseg)\n plt.imshow(seg_corner, cmap="Greys_r")\n plt.show()\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:23:05
get_ipython().magic(u'edit __')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n# plt.imshow(mseg, cmap="Greys_r")\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:23:13
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n# plt.imshow(mseg, cmap="Greys_r")\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:24:06
get_ipython().magic(u'reload_ext')
# nie, 10 lip 2016 18:24:50
get_ipython().magic(u'load_ext %autoreload')
# nie, 10 lip 2016 18:25:30
get_ipython().magic(u'edit _')
#[Out]# 'corners = np.zeros(img.shape, dtype=np.bool)\n\nfor i in range(0, nb_labels):\n mseg = img_label == i\n# plt.imshow(mseg, cmap="Greys_r")\n seg_corner = calc_shape_corners(mseg)\n corners = corners + seg_corner\n\n(n,m) = img.shape\nmat_x = np.tile(np.arange(0, m),n).reshape((n,m))\nmat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T\n\nxs = mat_x[corners]\nys = mat_y[corners]\n\nplt.imshow(img, cmap="Greys_r")\nplt.plot(xs, ys, "b+")\nplt.show()\n'
# nie, 10 lip 2016 18:26:10
with open("processing.py", "w+") as f:
f.write(Out[102])
# nie, 10 lip 2016 18:26:14
get_ipython().system(u'cat processing.py')

93
2015/2015_03/shoe-soles/old/try.py Normal file
View File

@ -0,0 +1,93 @@
# coding: utf-8
from dect import calc_shape_corners
calc_shape_corners(img)
from scipy import ndimage
img = ndimage.imread("/home/tpolgrabia/Pobrane/CSFID/tracks_cropped/00003.jpg")
img.shape
import matplotlib.pyplot as plt
plt.imshow(img, cmap="Greys_r")
plt.show()
from skimage import filters
otsu_lvl = filters.threshold_otsu(img)
otsu_lvl
plt.imshow(img, cmap="Greys_r")
plt.show()
plt.imshow(img >= otsu_lvl, cmap="Greys_r")
plt.show()
m = img <= otsu_lvl
plt.imshow(m, cmap="Greys_r")
plt.show()
img_label, nb_labels = ndimage.label(m)
nb_labels
plt.imshow(nb_labels)
plt.imshow(img_label))
plt.imshow(img_label)
plt.show()
sizes = ndimage.sum(m, img_label, range(0, nb_labels+1))
sizes
size_min_limit = 50
remove_segments = sizes < size_min_limit
remove_segments[img_label]
remove_mask = remove_segments[img_label]
img_label[remove_mask] = 0
plt.imshow(img_label)
plt.show()
import numpy as np
get_ipython().magic(u'pinfo np.unique')
labels = np.unique(img_label)
labels
get_ipython().magic(u'pinfo np.searchsorted')
img_label = np.searchsorted(labels, img_label)
nb_labels = len(labels)
plt.imshow(img_label)
plt.show()
ndimage.center_of_mass(img_label, labels, nb_labels+1)
ndimage.center_of_mass(img_label, labels, nb_labels+1)
get_ipython().magic(u'pinfo ndimage.center_of_mass')
ndimage.center_of_mass(m, img_label, labels)
centers = ndimage.center_of_mass(m, img_label, labels)
centers
centers[0]
centers[1]
centers[2]
centers[3]
centers[4]
centers[5]
labels]
labels
labels = np.unique(img_label)
labels
centers = ndimage.center_of_mass(m, img_label, labels)
centers
centers[:,0]
np.array(centers)
centers = np.array(centers)
centers[:,0]
xs = centers[:,0]
ys = centers[:,0]
xs = centers[:,1]
plt.imshow(m, cmap="Greys_r")
plt.plot(xs, ys, "r+")
plt.show()
get_ipython().magic(u'logstart')
get_ipython().magic(u'pwd ')
help %logstart
get_ipython().magic(u'pinfo %logstart')
get_ipython().magic(u'logstart -o')
get_ipython().magic(u'logstop')
get_ipython().magic(u'logstart -o')
get_ipython().magic(u'logstop')
get_ipython().magic(u'logstart -o -t')
get_ipython().magic(u'pwd ')
get_ipython().system(u'cat ipython_log.py')
pw
get_ipython().magic(u'edit')
get_ipython().magic(u'edit _')
get_ipython().magic(u'edit dect.py')
get_ipython().magic(u'edit __')
get_ipython().magic(u'edit _')
get_ipython().magic(u'edit _')
get_ipython().magic(u'edit __')
get_ipython().magic(u'edit __')
get_ipython().magic(u'edit _')
get_ipython().magic(u'edit test.py')

43
2015/2015_03/shoe-soles/poi_shoe_detection.py Normal file
View File

@ -0,0 +1,43 @@
# coding: utf-8
import sys
import matplotlib.pyplot as plt
from scipy import ndimage
path = sys.argv[1]
print("I am reading {}".format(path))
img = ndimage.imread(path)
print("I have read")
img_gray = 0.21 * img[:,:,0] + 0.72 * img[:,:,1] + 0.07 * img[:,:,2]
# plt.imshow(img_gray, cmap="Greys_r")
# plt.show()
from sklearn.cluster import KMeans
k = KMeans(n_clusters=5)
img_reshaped = img_gray.reshape((-1,1))
k.fit(img_reshaped)
values = k.cluster_centers_.squeeze()
labels = k.labels_
# labels
img_labels = labels.reshape(img_gray.shape)
# img_labels
# plt.imshow(img_labels)
# plt.show()
# get_ipython().magic(u'pinfo ndimage.median_filter')
# get_ipython().magic(u'pinfo ndimage.median_filter')
# get_ipython().magic(u'pinfo ndimage.median_filter')
img_labels2 = ndimage.median_filter(img_labels, 3)
# plt.imshow(img_labels2)
# plt.show()
from skimage.feature import corner_harris, corner_subpix, corner_peaks
coords = corner_peaks(corner_harris(img_labels2), min_distance=5)
coords_subpix = corner_subpix(img_labels2, coords, window_size=13)
fig, ax = plt.subplots()
ax.imshow(img_labels2, interpolation='nearest', cmap=plt.cm.gray)
ax.plot(coords[:, 1], coords[:, 0], '.b', markersize=3)
ax.plot(coords_subpix[:, 1], coords_subpix[:, 0], '+r', markersize=15)
# ax.axis((0, 350, 350, 0))
plt.show()
# fig, ax = plt.subplots()
# ax.imshow(img_labels2, interpolation='nearest', cmap=plt.cm.gray)
# ax.plot(coords[:, 1], coords[:, 0], '.b', markersize=3)
# ax.plot(coords_subpix[:, 1], coords_subpix[:, 0], '+r', markersize=15)
# plt.show()
# get_ipython().magic(u'save poi_shoe_detection.py 1-39')

242
2015/2015_03/shoe-soles/poi_shoe_detection2.py Normal file
View File

@ -0,0 +1,242 @@
import numpy as np
import math
from scipy import ndimage,spatial
import matplotlib.pyplot as plt
from skimage import feature
from skimage.feature import corner_harris, corner_subpix, corner_peaks
from sklearn.cluster import KMeans
from skimage.filters import threshold_otsu
from skimage.draw import line_aa
import math
import sys
def grayscale_luminosity(img):
return 0.21*img[:,:,0] + 0.72 * img[:,:,1] + 0.07 * img[:,:,2]
def kmeans_cluster(img_gray):
img_gray_reshaped = img_gray.reshape((-1,1))
print("About to start k-means clustering")
k = KMeans(n_clusters = 5)
print("Starting k-means clustring")
k.fit(img_gray_reshaped)
print("Finished k-means clustering")
values = k.cluster_centers_.squeeze()
labels = k.labels_
return (values, labels)
def corners_seg(img_labels2):
coords = corner_peaks(corner_harris(img_labels2), min_distance=5)
coords_subpix = corner_subpix(img_labels2, coords, window_size=13)
return (coords, coords_subpix)
def reassign_labels(labels1, values1, labels2, values2):
# TODO not efficient computionally
if values1.length != values2.length:
return None
n = values1.length
c_min = 1e8 # TODO max value of float
reassignment = np.array((n, 2))
for i in range(0, n):
e1 = values1[i]
for j in range(0, n):
e2 = values2[j]
d = np.sum(np.abs(e2-e1))
if d < c_min:
c_min = d
reassignment[i,0] = labels1[i]
reassignment[i,1] = labels2[j]
return reassignment
def compare_by_poi(img1, img2, xwindow = 5, ywindow = 5):
# both images are matrix of 0s or 1s
k = np.ones((xwindow,ywindow))
middlex = (xwindow-1) / 2
middley = (ywindow-1) / 2
k[middlex,middley] = 0
img2_neighbour_count_int = ndimage.convolve(img2)
img2_neighbour_count_bool = img2_neighbour_count_int > 0
img2_neighbour_count_int2 = np.array(img2_neighbour_count_bool, dtype=np.uint32)
img_comparision = img2_neighbour_count_int2 * img1
n1 = np.sum(img1)
n2 = np.sum(img_comparision)
return float(n2)/float(n1)
def find_first_marked_point(crossing_mask):
(n,m) = crossing_mask.shape
for i in range(0,n):
for j in range(0,m):
if crossing_mask[i,j]:
return (j,i)
return None
def MSE(a1, a2):
c = 1
for n in a1.shape:
c *= n
d = a1 - a2
return np.sum(d*d) / n
nsampling = 100
def calculate_shape_vector(
path,
nsampling,
median_filter_size = 10,
sweep_line_length = 1024.0):
img = ndimage.imread(path)
print("Image read")
img_gray_not_filtered = grayscale_luminosity(img)
img_gray = ndimage.median_filter(img_gray_not_filtered,
size=median_filter_size)
# fig = plt.figure(figsize=(15,15))
otsu_lvl = threshold_otsu(img_gray)
print(otsu_lvl)
mask = img_gray <= otsu_lvl
# plt.imshow(mask, cmap="Greys_r")
# plt.show()
otsu_sub_mask_lvl = threshold_otsu(img_gray[mask])
mask2 = img_gray <= otsu_sub_mask_lvl
sub_mask = mask * mask2
sub_mask2 = mask * (True ^ mask2)
# plt.imshow(sub_mask, cmap="Greys_r")
# plt.show()
# plt.imshow(sub_mask2, cmap="Greys_r")
# plt.show()
label_im, nb_labels = ndimage.label(sub_mask2)
print("Nr of regions: {}".format(nb_labels))
sizes = ndimage.sum(sub_mask2, label_im, range(nb_labels + 1))
mask_size = sizes < 500
print("Liczba usunietych: {}".format(np.sum(mask_size)))
remove_pixel = mask_size[label_im]
label_im[remove_pixel] = 0
# plt.imshow(label_im)
# plt.show()
centers = np.zeros((nb_labels+1-np.sum(mask_size),2))
idx = 0
for i in range(0, nb_labels+1):
if mask_size[i]:
continue
# processing that means calculating segment parameters
mask_segment = label_im == i
centers[idx,:] = ndimage.center_of_mass(mask_segment)
idx += 1
print("Centers: {}".format(centers))
centers2 = np.copy(centers)
(n,m) = img_gray.shape
centers2[:,0] = centers[:,0] / float(n)
centers2[:,1] = centers[:,0] / float(m)
print("Centers 2: {}".format(centers2))
mask_int = np.array(mask, dtype=np.uint32)
label_im, nb_labels = ndimage.label(mask_int)
sizes = ndimage.sum(mask, label_im, range(nb_labels + 1))
mask_size = sizes < 0
remove_pixel = mask_size[label_im]
label_im[remove_pixel] = 0
edges2 = feature.canny(label_im > 0)
# plt.imshow(edges2, cmap="Greys_r")
sx = 0.0
sy = 0.0
c = 0
(n,m) = edges2.shape
print("Width: {}, height: {}".format(n, m))
for i in range(0, n):
for j in range(0, m):
if edges2[i,j]:
c += 1
sx += j
sy += i
sx /= c
sy /= c
print("Middle ({}x{})".format(sx,sy))
# plt.plot([sx], [sy], "r+")
line_length = sweep_line_length
angle_step = 2.0 * np.pi / nsampling
dist_vector = np.zeros((nsampling), dtype=np.float32)
dist_vector[:] = 1e8 # TODO add here infinitium
for i in range(0,nsampling):
line_arr = np.zeros((n,m), dtype=np.bool)
rr,cc,val = line_aa(int(math.floor(sy)),
int(math.floor(sx)),
int(math.floor(sy+line_length*math.sin(angle_step*i))),
int(math.floor(sx+line_length*math.cos(angle_step*i))))
m1 = rr >= 0
m2 = rr < n
m3 = cc >= 0
m4 = cc < m
mall = m1 * m2 * m3 * m4
cc2 = cc[mall]
rr2 = rr[mall]
val2 = val[mall]
line_arr[rr2,cc2] = val2 > 0
crossing_mask = (edges2 * line_arr)
# TODO find one (first - best) crossing point and mark him
p = find_first_marked_point(crossing_mask)
if p != None:
plt.plot([p[0]], [p[1]], "r+")
dist_vector[i] = math.sqrt((p[0] - sx)*(p[0] - sx) + (p[1] - sy)*(p[1] - sy))
# plt.show()
return (mask, dist_vector / np.min(dist_vector), centers2)
if len(sys.argv) < 2:
print("Too small arguments")
sys.exit(1)
(img_gray1, v1, centers1) = calculate_shape_vector(sys.argv[1],nsampling)
if len(sys.argv) < 3:
print("To small arguments to compare")
sys.exit(0)
(img_gray2, v2, centers2) = calculate_shape_vector(sys.argv[2],nsampling)
# (img_gray2, v2) = calculate_shape_vector("italian-sole-new-800_1_2_1.jpg",nsampling)
def compare_sets(c1, c2):
k = spatial.KDTree(c1)
s = 0.0
(n,m) = c2.shape
for i in range(0, n):
(d,idx) = k.query(c2[i,:])
s += d*d
s /= float(n)
return math.sqrt(s)
print("v1")
print(v1)
print("v2")
print(v2)
diff = v1-v2
print("Diff: {}".format(diff))
mse = np.sum(diff*diff)/nsampling
print("MSE: {}".format(MSE(v1,v2)))
print("Diff between sets of points: {}"
.format(compare_sets(centers1, centers2)))

73
2015/2015_03/shoe-soles/print-processing.py Normal file
View File

@ -0,0 +1,73 @@
from scipy import ndimage
import math
import matplotlib.pyplot as plt
from skimage import filters
import numpy as np
from skimage import feature
img = ndimage.imread("/home/tpolgrabia/Pobrane/CSFID/tracks_cropped/00003.jpg")
otsu_lvl = filters.threshold_otsu(img)
img_mask = img >= otsu_lvl
img_mask_fil = True ^ ndimage.minimum_filter(img_mask, 3)
img_label, nb_labels = ndimage.label(img_mask_fil)
sizes = ndimage.sum(img_mask_fil, img_label, range(0, nb_labels+1))
remove_labels = sizes < 100
lbls = True & remove_labels
lbls[0] = False
print("Remove labels: {}".format(ndimage.sum(remove_labels)))
remove_mask = remove_labels[img_label]
img_label[remove_mask] = 0
idx = 0
n2 = nb_labels + 1 - np.sum(remove_labels)
centers = np.zeros((nb_labels+1,2))
(n,m) = img.shape
corners = np.zeros((n,m), dtype=np.bool)
for i in range(0, nb_labels+1):
if remove_labels[i]:
continue
seg = img_label == i
slice_y, slice_x = ndimage.find_objects(seg)[0]
seg_window = seg[slice_y, slice_x]
edge = feature.canny(seg_window)
plt.imshow(edge, cmap="Greys_r")
plt.show()
(en,em) = edge.shape
distances = np.zeros((en,em))
centers[idx,:] = ndimage.center_of_mass(seg)
for j in range(0, en):
for k in range(0, em):
if edge[j,k]:
diff = centers[i,:] - np.array([slice_y.start, slice_x.start])
dx = k - diff[1]
dy = j - diff[0]
d = math.sqrt(dx*dx+dy*dy)
distances[j,k] = d
# max_distances = ndimage.maximum_filter(distances, 20)
# t_corners = np.abs(max_distances - distances) < 1e-6
# print(t_corners)
# plt.imshow(t_corners, cmap="Greys_r")
# plt.show()
# corners[slice_y,slice_x] = corners[slice_y,slice_x] + edge
idx += 1
(n,m) = img.shape
mx = np.tile(np.arange(0,m),n).reshape((n,m))
my = np.tile(np.arange(0,n),m).reshape((m,n)).T
xs2 = mx[corners]
ys2 = my[corners]
xs = centers[:,1]
ys = centers[:,0]
plt.imshow(img_label, cmap="Greys_r")
# plt.plot(xs, ys, "r+")
plt.plot(xs2,ys2,"b+")
plt.show()

63
2015/2015_03/shoe-soles/processing.py Normal file
View File

@ -0,0 +1,63 @@
from dect import calc_shape_corners
import math
from scipy import ndimage
from skimage import filters
import numpy as np
import matplotlib.pyplot as plt
img = ndimage.imread("/home/tpolgrabia/Pobrane/CSFID/tracks_cropped/00003.jpg")
(n,m) = img.shape
otsu_lvl = filters.threshold_otsu(img)
m = img <= otsu_lvl
img_label, nb_labels = ndimage.label(m)
sizes = ndimage.sum(m, img_label, range(0, nb_labels+1))
min_seg_size = 100
to_be_removed_segs = sizes < min_seg_size
remove_mask = to_be_removed_segs[img_label]
img_label[remove_mask] = 0
labels = np.unique(img_label)
nb_labels = len(labels)
img_label = np.searchsorted(labels, img_label)
labels = np.unique(img_label)
centers = ndimage.center_of_mass(m, img_label, labels)
centers = np.array(centers, dtype=np.float)
(nc,who_cares) = centers.shape
def distance_point(m1, m2):
df = m1-m2
df2 = df*df
sd2 = np.sum(df2)
d = math.sqrt(sd2)
return d
# TODO check if generates all rates of distance pairs
for i in range(0, nc):
for j in range(0, i):
for k in range(0, j):
m1 = centers[i,:]
m2 = centers[j,:]
m3 = centers[k,:]
d1 = distance_point(m1, m2)
d2 = distance_point(m2, m3)
d = d1 / d2
print("{}-{}/{}-{}: {}".format(i, j, j, k, d))
corners = np.zeros(img.shape, dtype=np.bool)
for i in range(0, nb_labels):
mseg = img_label == i
# plt.imshow(mseg, cmap="Greys_r")
seg_corner = calc_shape_corners(mseg)
corners = corners + seg_corner
(n,m) = img.shape
mat_x = np.tile(np.arange(0, m),n).reshape((n,m))
mat_y = np.tile(np.arange(0, n),m).reshape((m,n)).T
xs = mat_x[corners]
ys = mat_y[corners]
plt.imshow(img, cmap="Greys_r")
plt.plot(xs, ys, "b+")
plt.show()

114
2015/2015_03/shoe-soles/sole-detection.py Normal file
View File

@ -0,0 +1,114 @@
#!/usr/bin/python3
import numpy as np
import matplotlib.pyplot as plt
import sys
from scipy import ndimage
from skimage import feature
from skimage import filters
from skimage.feature import corner_harris, corner_subpix, corner_peaks
def detect_corner(img, windowx = 2, windowy = 2):
(n,m) = img.shape
corners = np.zeros((n,m), dtype=np.bool)
nr_cuts = 0
for i in range(0, n):
print("Row: {}".format(i))
for j in range(0, m):
corner = False
for k in range(-windowy, windowy):
nr_cuts = 0
for l in range(-windowx, windowx):
y = (i+k)%n
x1 = (j+l)%m
x2 = (j+l+1)%m
if img[y,x1] ^ img[y,x2]:
nr_cuts += 1
if nr_cuts > 1:
corner = True
# print("{}x{} is corner".format(j,i))
break
corners[i,j] = corner
return corners
def produce_poi_list(img):
r = []
(n,m) = img.shape
for i in range(0, n):
row = []
for j in range(0, m):
if img[i,j]:
row.append((float(j)/m,float(i)/n))
r.append(row)
return r
# not time-effective comparision
# TODO compare only rows in the r window
def compare_poi_list(pl1, pl2, r):
matches = []
for pr1 in pl1:
for (x1,y1) in pr1:
for pr2 in pl2:
for (x2,y2) in pr2:
if r * r > ((x1 - x2)*(x1 - x2) + (y1-y2)*(y1-y2)):
matches.append(((x1,y1),(x2,y2)))
return matches
file_path = sys.argv[1]
print("File {} to be analysed".format(file_path))
img = ndimage.imread(file_path)
img_gray = 0.21 * img[:,:,0] + 0.72 * img[:,:,1] + 0.07 * img[:,:,2]
n,m = img_gray.shape
print("Width {}, Height {}".format(m,n))
otsu_lvl = filters.threshold_otsu(img_gray)
seg_shoe = img_gray <= otsu_lvl
img_shoe_selected = seg_shoe * img_gray
data_selected = img_gray[seg_shoe]
shoe_otsu_lvl = filters.threshold_otsu(data_selected)
img_sole_segs1 = img_shoe_selected <= shoe_otsu_lvl
img_sole_segs = ndimage.median_filter(img_sole_segs1, 5)
img_sole_segs_nr = np.array(img_sole_segs, np.int)
k = np.ones((11, 11))
img_sole_seg_count = ndimage.convolve(img_sole_segs_nr, k)
m1 = img_sole_seg_count >= 80
m2 = True ^ img_sole_segs
# m2 = img_sole_seg_count <= 1
m3 = m1 * m2
# m3i = np.array(m3, np.int)
corners = detect_corner(img_sole_segs)
poi_list = produce_poi_list(corners)
print(poi_list)
m3i = np.array(corners, np.int)
# color_red = np.ones((n,m,3)) * np.array([255, 0, 0])
color_red = 255
img_annotated = img_gray * (1 - m3i) + m3i * color_red
# img_annotated = img_gray
# plt.imshow(m1 * m2, cmap="Greys_r")
# plt.imshow(img_annotated, cmap="Greys_r")
# coords = corner_peaks(corner_harris(img_gray), min_distance=5)
# coords_subpix = corner_subpix(img_gray, coords, window_size=13)
fig = plt.figure()
ax1 = fig.add_subplot(2,1,1)
ax1.imshow(img_sole_segs, cmap="Greys_r")
ax2 = fig.add_subplot(2,1,2)
ax2.imshow(img_annotated, cmap="Greys_r")
# ax2.plot(coords[:, 1], coords[:, 0], '.b', markersize=3)
# ax2.plot(coords_subpix[:, 1], coords_subpix[:, 0], '+r', markersize=15)
plt.show()
plt.imsave("sole_seg.png", img_sole_segs)