Multiclass object recognition inspired by biological vision

The problem of recognizing multiple object classes in natural images has proven to be a difficult challenge for compute vision. It is reasonable to look to biology for inspiration, a novel multiclass object recognition algorithm based on a biologically inspired model named ST model is proposed. ST model is based on the theory of biological neurology, which calculates object features that exhibit position and scale invariance with uniform model, whose performance on object recognition is superexcellent. Firstly, the standard model features of object are extracted via ST model. And then, the identification matrix of multiclass object is built on the need of recognition. At last, the task of recognizing multiclass object is successfully completed with SVM. Experimental results show that our approach exhibits excellent recognition performance.