Abstrato

The shape recognition of agricultural products and invariance analysis based on simulation of optical correlation recognition

Min Hao, Zheying Zong, Jianli Cui, Zhihong Yu, Haiqing Tian

At present, a method of computer simulation of the optical joint transform correlation is widely used in the fingerprint identification but rarely in the agricultural products detection. This paper introduces theory of the optical correlation recognition,simulates by Fourier Transform and choose the maximum sidelobe as the feature to recognize the shape of agricultural products. Take the potato for example,the paper test the feasibility of the algorithm and verify the invariance. This algorithm is easy and feasible, has the invariance of translation, scaling and rotating(a small azimuth angle range). With low requirement of image and rapid calculation,this method can be widely applied in the shape recognition of many agricultural products.