OpenCV视觉分类库

Project Description This library provides a flexible feature-based image classification/object recognition framework made to augment OpenCV. It is effective at classifying objects into broad categories (as demonstrated by the terrain classification demo) as well as very unique objects. Classification is performed using the Bag of Visual Words (BOVW) data structure with linear Support Vector Machines (SVMs). SURF and SIFT features are supported and the color histogram-based classification is also included (primarily for comparison). The framework allows a large degree of customization through XML setup files. Users can easily create and classify large image databases. Project Demo height=344 />