This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
Classifying single surf descriptors is not the way to go, the easiest way is to create global descriptors. Either you form a bag-of-words descriptor from local descriptors, see e.g. http://answers.opencv.org/question/8677/image-comparison-with-a-database/#8686, or you use global descriptors like GIST (google for it). For these global descriptors you can then train a classifier (e.g. SVM). You can also use the cascade-classifier of OpenCV, see http://docs.opencv.org/doc/user_guide/ug_traincascade.html. Here also a global descriptor (HOG/LBP/HAAR) is formed and then classified using a cascade of simple classifiers.
Please also note that this question has been asked and answered a lot, so please use the search function!
