Meat product freshness detection method and system

The invention relates to the technical field of chicken detection, in particular to a meat product freshness detection method and system.The method comprises the steps that chicken images of different freshness classifications are collected, and data sets are divided; the method comprises the steps that an improved CNN network is constructed, the improved CNN network is composed of six layers of convolutional neural networks and an average pooling layer, after average pooling, Dropout is utilized to randomly discard part of neurons according to a set proportion, a leakyrelu activation function is adopted, and a BN layer is inserted in front of each activation layer; and finally, accessing a full connection layer and a softmax classifier. According to the method, an improved CNN network is constructed, six layers of convolution structures are utilized, Inception modules are adopted in the fifth and sixth layers of convolution modules, neurons of receptive fields of different sizes are activated, texture recogn