CODEWITHSUNDEEP
pythonopencv
Baqir Khan
Baqir Khan

Reputation: 804

Efficient way to know if any pixel is white in image?

So basically I m trying to use OpenCV-Python to do motion detection. I used this tutorial to do so and here is my code.

import cv2

def diffImg(t0, t1, t2):
    d1 = cv2.absdiff(t2, t1)
    d2 = cv2.absdiff(t1, t0)
    return cv2.bitwise_and(d1, d2)


cap = cv2.VideoCapture(0)

t = cv2.cvtColor(cap.read()[1], cv2.COLOR_BGR2GRAY)
tp = cv2.cvtColor(cap.read()[1], cv2.COLOR_BGR2GRAY)
tpp = cv2.cvtColor(cap.read()[1], cv2.COLOR_BGR2GRAY)

while cap.isOpened():
    img = cv2.cvtColor(cap.read()[1], cv2.COLOR_BGR2GRAY)
    img2 = diffImg(t,tp,tpp)

cv2.imshow("Motion", img2)
t=tp
tp=tpp
tpp=img

key = cv2.waitKey(10)

if key == 27 :
    cv2.destroyAllWindows()
    break

The I want to print on the console when there is a motion detection or not. When there is a motion, there are white pixels in the input image. But I don't know how to find white pixels in the input image. Can anyone tell me how to find if there are white pixels in the image returned by diffImg or not ?

Upvotes: 2

Views: 571

Answers (1)

David Gasquez
David Gasquez

Reputation: 327

You could take a look to the countNonZero function fom OpenCV.

Example provided by Baqir Khan: 

if cv2.countNonZero(img2) > 29700: 
    print("Motion") 
else: 
    print("No Motion")

Upvotes: 1

Related Questions