I'm working on images , where I intend to extract glcm, lbp , morphological features and also use a CNN network to get CNN based feature map. Is there any network or strategy where I can feed the image features and CNN extracted features to a network to perform feature selection?
I read papers where they use multi source data for feature extraction , but the data they used were images in both the sources. I could have tensors, matrix , image ... as input to a network that selects the best features from all feature extraction techniques.
