CODEWITHSUNDEEP
pythonimage-processingdeep-learningcanny-operator
user34
user34

Reputation: 46

How to differentiate between two types of images using image processing?

I have two categories of images segmented using MaskRCNN. First category contains dent images on car.

Second category contains images of reflection/shadows fallen on car that the Mask RCNN detects as dent.

Is there any methods in image processing that can distinguish between the same? I tried Canny, Gaussian, LBP, SIFT, watershed etc. Can anyone suggest a suitable approach for the same.

Thanks in advance!!

Upvotes: 0

Views: 205

Answers (1)

Matt Yoon
Matt Yoon

Reputation: 446

If the image of shadows generally are darker than the dents, you can convert the images to HSV, and compare the average V values of the pixels. V value contains the brightness of the pixel.

You can use hsv_img = cv2.cvtColor(img, cv2.BGR2HSV).

Upvotes: 1

Related Questions