mirror of
https://github.com/opencv/opencv.git
synced 2024-11-28 13:10:12 +08:00
67 lines
2.1 KiB
Python
Executable File
67 lines
2.1 KiB
Python
Executable File
#!/usr/bin/python
|
|
from opencv.cv import *
|
|
from opencv.highgui import *
|
|
from random import randint
|
|
MAX_CLUSTERS = 5
|
|
|
|
if __name__ == "__main__":
|
|
|
|
color_tab = [
|
|
CV_RGB(255,0,0),
|
|
CV_RGB(0,255,0),
|
|
CV_RGB(100,100,255),
|
|
CV_RGB(255,0,255),
|
|
CV_RGB(255,255,0)]
|
|
img = cvCreateImage( cvSize( 500, 500 ), 8, 3 )
|
|
rng = cvRNG(-1)
|
|
|
|
cvNamedWindow( "clusters", 1 )
|
|
|
|
while True:
|
|
cluster_count = randint(2, MAX_CLUSTERS)
|
|
sample_count = randint(1, 1000)
|
|
points = cvCreateMat( sample_count, 1, CV_32FC2 )
|
|
clusters = cvCreateMat( sample_count, 1, CV_32SC1 )
|
|
|
|
# generate random sample from multigaussian distribution
|
|
for k in range(cluster_count):
|
|
center = CvPoint()
|
|
center.x = cvRandInt(rng)%img.width
|
|
center.y = cvRandInt(rng)%img.height
|
|
first = k*sample_count/cluster_count
|
|
last = sample_count
|
|
if k != cluster_count:
|
|
last = (k+1)*sample_count/cluster_count
|
|
|
|
point_chunk = cvGetRows(points, first, last)
|
|
|
|
cvRandArr( rng, point_chunk, CV_RAND_NORMAL,
|
|
cvScalar(center.x,center.y,0,0),
|
|
cvScalar(img.width*0.1,img.height*0.1,0,0))
|
|
|
|
|
|
# shuffle samples
|
|
cvRandShuffle( points, rng )
|
|
|
|
cvKMeans2( points, cluster_count, clusters,
|
|
cvTermCriteria( CV_TERMCRIT_EPS+CV_TERMCRIT_ITER, 10, 1.0 ))
|
|
|
|
cvZero( img )
|
|
|
|
for i in range(sample_count):
|
|
cluster_idx = clusters[i]
|
|
# a multi channel matrix access returns a scalar of
|
|
#dimension 4,0, which is not considerate a cvPoint
|
|
#we have to create a tuple with the first two elements
|
|
pt = (cvRound(points[i][0]), cvRound(points[i][1]))
|
|
cvCircle( img, pt, 2, color_tab[cluster_idx], CV_FILLED, CV_AA, 0 )
|
|
|
|
cvShowImage( "clusters", img )
|
|
|
|
key = cvWaitKey(0)
|
|
if( key == 27 or key == 'q' or key == 'Q' ): # 'ESC'
|
|
break
|
|
|
|
|
|
cvDestroyWindow( "clusters" )
|