How should I approach preparation of training data/cascade training with one object which can be one of two colors?
I have to detect one object which exists in a blue and red state. Should I use separate positive samples and do separate cascade training for each object, or can one cascade classifier be used to identify the object and determine if it is the red or blue form? I'd imagineit'd be much more efficient to use one cascade classifier, if this is possible. Thanks in advance
are you aware, that cascades only look at grayscale images ? ("shape" detectors ?)
(most probably, this is not a solution to your problem at all !)