Pork freshness pattern recognition based on SOM neural network

摘要

In this paper, using digital image processing techniques to detect plaque bacteria that taking hough transform to extract complete outline of fat cell based on mathematical morphology method. TVB-N values in the process of pork Corruption is taken for a sequence of inputs of SOM neural network used for the cluster analysis. CCD photoelectric detection techniques that can be characterized by a number of fresh pork in the amount of testing. The final study of using neural network technology multi-data fusion detection methods, in order to achieve the freshness of the meat inspection classification identification.