The invention provides a tire code and tire state identification method based on
machine learning, comprising the following steps: S1, acquiring a tire image, and preprocessing the image; S2, establishing a single-layer neural
network model, and calculating a model output function; S3, based on an error
back propagation algorithm, iteratively updating S2, and establishing a multi-layer neural
network model which comprises an input layer, a
hidden layer and an output layer; S4, establishing a
convolutional neural network as a
hidden layer in the multi-layer neural network; S5, establishing a full connection layer to connect the
hidden layer with the output layer, and outputting the tire code and the tire state of the tire by the output layer; S6, carrying out supervising learning on the multilayer neural network through the artificially marked
verification set, and updating neural network parameters; and S7,
processing the tire image by using the trained multi-layer neural network to obtain a corresponding tire code and a tire state. According to the method, the tire
inspection time is effectively shortened, the identification precision is high, and convenience is provided for
automobile safety inspection, accurate
tire maintenance, tire leasing service and the like.