A function which, given an image, returns a response for every cascade classifier by the output of #59.
The natural response given by CascadeClassifier is location and number of occurrences of the the search pattern. For us, this will essentially be a binary response as we will only be searching on a scale appropriate to find the pattern at the whole-image scale. However, we could possibly use the numDetections response (the number of positively identified rectangles merged to make the detection) as a scalar output instead, which would give us more information to work with.
This should be done in Python so it is able to interface with a downstream sklearn classification from which we can get a probability for of the image belonging to each of the classes.
A function which, given an image, returns a response for every cascade classifier by the output of #59.
The natural response given by CascadeClassifier is location and number of occurrences of the the search pattern. For us, this will essentially be a binary response as we will only be searching on a scale appropriate to find the pattern at the whole-image scale. However, we could possibly use the numDetections response (the number of positively identified rectangles merged to make the detection) as a scalar output instead, which would give us more information to work with.
This should be done in Python so it is able to interface with a downstream sklearn classification from which we can get a probability for of the image belonging to each of the classes.
http://docs.opencv.org/trunk/doc/py_tutorials/py_objdetect/py_face_detection/py_face_detection.html https://github.com/Itseez/opencv/blob/master/samples/python2/facedetect.py http://docs.opencv.org/trunk/modules/objdetect/doc/cascade_classification.html