I am doing machine learning for image classification, now I am moving to object detection. I need to segment a whole image and pass the resulting blob to my classifiers. I know about SSD,Yolo and faster R-CNN but my approach is simpler
Hi, you might want to try scikit-image (http://scikit-image.org/), it offers most classical image processing operations. Following your segmentation, the characteristics of labeled region can be measured using the "measure" submodule.