Neural network model firmware generation method and image recognition method

The invention relates to a model firmware generation method of a neural network and an image recognition method in the technical field of image processing, and the method comprises the following steps: configuring a plurality of groups of different input resolutions as the input of the neural network, and analyzing the neural network to obtain model parameters; generating a quantized picture set, executing a forward process on the quantized picture set to obtain forward quantized data of each layer of neural network, and obtaining quantized parameter data and fixed-point weight data; calculating a first running memory, occupied by each layer of neural network, of the embedded board end under each group of input resolutions to obtain a first memory occupancy set; configuring a plurality of groups of neural networks, and calculating a second running memory, occupied by each group of neural networks, of the embedded board end under each group of input resolutions to obtain a second memory occupancy set; the meth