I want to perform object recognition using images and videos but i dont know where to start from?

edit

I am recently working on object detection using traincascade in opencv which i have learned a lot and prepared an .xml file to detect tomato but I have to perform object recognition . Please suggest me what to start? how to start? what to learn? how to implement? Please help me asap.