In recent years, due to the rapid development of computer technology, artificial intelligence technology in the computer field has begun to integrate into people’s life, and facial recognition, as a unique biometric recognition method, is the core of artificial intelligence technology. Based on this, this paper discusses the local feature extraction and global feature extraction based on the deep learning algorithm, and proposes a training classification method based on the deep learning model combined with local pattern and GLQP representation feature extraction algorithm. In this paper, the local quantization method is used to input the data set preprocessed by the filter into the network. The depth of CNN network is selected as 4 layers, and the network is trained to produce high-resolution features. Experiments show that the accuracy of the trained deep network model is 92.2% in the test set. Therefore, compared with the traditional methods, deep learning has the advantages of powerful visualization and automatic face feature extraction, overcomes the shortcomings of deep learning model in the process of shallow feature learning, and shows higher recognition efficiency and generalization.

Super resolution image reconstruction technology is the core technology in image processing, which provides technical support for most of the intelligent devices for target tracking and target detection. However, the existing image acquisition stage will cause extra high cost to improve the image resolution. Optimization algorithm is one of the best ways to solve this kind of problem. At present, there is no substantial breakthrough in this field in China. Therefore, this paper studies single image super-resolution reconstruction algorithm based on deep learning. In this paper, the application of super-resolution reconstruction algorithm in single image is discussed. The analysis shows that the application of deep learning in this field is shallow and has a large optimization space. Especially for the existing main problems, the traditional algorithm is optimized and improved. According to the demand of resolution reconstruction, combined with the latest deep learning method, the accuracy and robustness of the algorithm are further improved. At the same time, it effectively improves the comprehensive performance of the model. In order to further verify the actual performance of the proposed algorithm, the corresponding comparative experiments are established. The experimental results show that the proposed algorithm has obvious advantages over the traditional fsrcnn algorithm, especially when the PSNR of dataset 4 is 2, 3, and 4 times, it can increase 0.41 dB, 0.58 dB and 0.52 DB respectively. Analysis shows that this algorithm has obvious advantages and achieves ideal results.