Food Image Recognition Algorithm Base on Improved VGG16

Being aimed at the problems of difficult feature extraction and low recognition accuracy of oven food image in the existing network model, the improved VGG16 oven food classification method is proposed. This method uses Asymmetric Convolution Block (ACB) to replace the original convolution kernel to improve the performance of the benchmark model and add a Batch Normalization (BN) layer after the pooling layer to enhance standardization. However, due to the problems of high texture similarity, complex background and background interference in oven food image, the attention mechanism is added to the convolution neural network model. Experimental results prove that compared with the traditional neural network the recognition accuracy of this model is improved to 99.2%. The model proposed in this paper can effectively improve the accuracy and applicability of image classification.