Distinguish between the data of the Internet of Things based on abnormal audio detection and the image of athletes’ human actions

This article discusses methods of abnormal audio detection. Extract audio information from audio acquisition hardware, convert analog signals into digital signals, perform pre-focused audio processing on digital signals, and compare and analyze audio noise reduction algorithms, such as wave noise reduction, homomorphic filtering, and Wiener filtering. After noise reduction processing is performed on the audio data, endpoint detection is performed on the continuous noise reduction frequency data, and the actual audio segment is separated from the continuous audio signal. Consistent with this, the blockchain-based IoT data service reputation evaluation model designed and implemented in this paper can support the data service reputation evaluation requirements. Users can upload IoT data sources to the platform or retrieve data from the platform, and research the data based on reputation or data provider analysis. The interaction with the data authority is realized through access control based on permissions. Finally, according to the latest requirements of the national sports team, based on the research foundation of the “athlete military division,” a new visual word model is used to explain the research on the image discrimination of athletes’ movements, and introduces the athletes’ movements, including traditional word models and automatic algorithms, and then introduces the research content and research methods to improve the traditional new visual word model to adapt to the characteristics of athletes. This algorithm minimizes the impact of short recognition rate and slow recognition rate, the latter causing problems such as the background similarity of the athlete’s image and the difference between the athlete’s actions. By understanding the available modules in the athlete’s motion recognition system, the prototype design of the basic functions and the athlete’s video system was completed.