Artificial neural network normalization system for identification device

The invention discloses an artificial neural network normalization system for an identification device, and the system comprises an input layer which generates an initial feature map of an image; a plurality of hidden layers for convolution of the initial feature map to generate an object feature map; and a matching unit is used for receiving the object feature map and executing matching so as tooutput an identification result. The first deduction block and the second deduction block are arranged in at least one hidden layer of the artificial neural network. In the first mode, the first deduction block is opened and the second deduction block is closed, and the first deduction block only receives the output of the first deduction block of the previous layer. In the second mode, the firstdeduction block and the second deduction block are started, and the second deduction block receives the output of the second deduction block of the previous layer and the output of the first deductionblock of the previous laye