Development of image recognition based on artificial intelligence

Abstract

With the development of computer science, electronics and artificial intelligence technology, image detection technology is becoming more and more mature, and the application of object recognition and detection through image detection technology is becoming more and more extensive. This paper mainly studies the development of image recognition based on artificial intelligence technology. This paper first analyzes the design method of classical convolutional network, hoping to achieve the accuracy level of similar methods in small-scale image recognition with more simplified network parameters and model scale, or even surpass the same methods. In this paper, a multi-structured convolutional neural network is designed for image recognition. The experimental results show that the DSCNN model designed in this paper has higher image recognition accuracy.