Image Semantic Analysis Based On Structure Of Optimization Deep Neural Network

With the rapid development of the Internet,the data of image appears in people's lives frequently,and grow up quickly.The research of image semantic analysis has been a hot topic in the field of computer vision.Many researchers have proposed a variety of methods to learn the underlying features of the image,but the semantic gap between the low-level visual features and high-level semantics of the image remains to be solved.In recent years,deep learning is emerging in the field of computer vision research.It has been successfully applied to the retrieval of image data by exploring the advanced neural network to optimize the image features.Therefore,this paper proposes an image retrieval algorithm based on deep neural network structure optimization.The method of this paper is divided into three steps: firstly,preprocessing the image data set to make the size of the image is the same.Secondly,in the process of training model of convolutional neural network,optimizes the network structure,improve the convergence speed of the network,reducing the network model parameters extraction,while using the trained model for image feature data set.Finally,the extracted features are combined with the corresponding similarity measurement algorithm to get the final retrieval results.It can be seen from the comparison of the experimental results that the algorithm proposed in this paper is superior and effective.At the same time,using the evaluation standard of image retrieval,such as precision,recall.Comparing with traditional feature extraction method and similarity measure methods,the experimental results show that our algorithm has higher accuracy.