Computer Vision and Image Processing Under Big Data Analysis

Currently, cutting-edge technologies such as computer technology and network technology are rapidly developing, and they have been widely applied in various industries and fields. At the same time, a large amount of data information is constantly being generated. Therefore, it is necessary to strengthen, optimize and analyze various disorderly and fragmented information to ensure that these disorderly information can be effectively removed, thereby improving the reliability and accuracy of information data. Compared to traditional BP neural networks, this paper proposes a new image processing method based on CNN (Convolutional Neural Networks). The article combines the concepts and technical points of computer image processing technology and big data analysis and recognition technology, analyzes their advantages and principles, and discusses the specific application strategies of this technology. In such a large environment, utilizing computer image processing and recognition technology can better utilize its flexibility and processing capabilities, and enhance its application in various industries. Finally, this paper verifies the progressiveness of the tennis error action recognition model based on CNN algorithm through experiments (with the increase of the number of tennis error techniques, the recall rate of the tennis action recognition model based on CNN algorithm continues to increase to 89%, higher than other algorithms).