Category "haar-classifier"

Face detection return value without face

I' using haar cascade for face detection faces_haar = face_cascade.detectMultiScale(image, scaleFactor=1.3, minNeighbors=4, minSize=(30, 30), flags=cv2.CASCADE

opencv_traincascade -numStages 1 How many stages do I need?

I am tracking one particular dog and nothing else. What is the maximum stage, number of positive and negative samples that makes sense? With: ./opencv_trainc

Other Categories