How to compute tresholded (black) pixels after binarization?
Here is a Python code:
img = cv2.imread('imageexperiment.tif',0)
imageone = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
Now i want somehow to compute number of black pixels (stained area)
You want to compute the number of pixels not null, there is a function for that: [countNonZero] (http://docs.opencv.org/modules/core/d...)
That's because cv2.threshold returns 2 arguments (see the doc). Change your line with
ret, imageone = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
arearesult = cv2.countNonZero(imageone)
and it should work fine.
@Mathieu: Now I am confused: what is the retval?
@Mat29: If you want to have the number of black pixels, you can subtract the arearesult from the imagearea; because the countNonZero funciton is returning (as expected) the number of non-zero pixels (that is the number of white pixels)
You're right, you need to subtract the image size to the number of non zero pixels. retval is indeed undocumented, but It seem it is the threshold value.
I suggest to make another question for that and add the code, too. Or just test it and tell us what have you done and if it has changed ;)
Asked: 2015-06-08 08:06:55 -0600
Seen: 101 times
Last updated: Jun 08 '15
imageone .rows*imageone .cols-sum(imageone )
This question has already been answered more than once...