An Improved Method of Detecting Pork Freshness Based on CRR Features

摘要

In this paper, an improved pork freshness detecting method based on color features was presented according to the biochemical mechanism of pork metamorphism. The pork sample images were captured and processed by de-noisy and segmentation methods. Then, the fat areas were eliminated to reduce the influence of detection in HSV color space. A new Color Region Ratio (CRR) extraction method was presented by red ratio. 9 groups sample data were classified to conclude the principle of detecting pork freshness and computed the correlation coefficients. Finally, the PNN algorithm in neural network was performed to detect pork freshness. The experimental results showed that the color features were correlated with freshness and could detect it effectively.