In the current information technology, the basic feature of dynamic image recognition is to digitize the biological information of different individuals. Compare the features that can be identified, and then realize the recognition function. This article first investigates the current status of the current clothing sales platform management method, introduces the main situation of AI learning, analyzes the work flow of the sales platform, and studies the current needs, and then focuses on whether this technology can improve the platform The work efficiency of the company launched an in-depth discussion. Secondly, it analyzes some problems that may exist in clothing sales management, and then points out a new recognition algorithm based on these analysis results. This algorithm is a combination of dynamic image analysis and deep neural network, which can be used for clothing such as factories. The video provided by the unit conducts dynamic analysis, automatically recognizes and organizes clothing online design and sales, realizes efficient and intelligent clothing transaction statistics, and this makes the whole system run much more efficiently. The design of the system is combined with various problems existing in the process of clothing marketing, and various applications of the system in clothing design and marketing are discussed, so as to give full play to the advantages of dynamic image recognition technology and AI intelligence technology. At the same time, the implementation plan is also expressed, and part of the interface is explained. Finally, this article describes the test situation of the system, mainly including the function test and performance test of the project.