Research And Application Of Human Detection Technology

As a scientific and technological means to promote social progress and development,the Internet and the Internet of Things are receiving more and more attention from the state and society.With the gradual promotion of the construction of"Safe Campus" and "Smart Campus",it has become an urgent need to scientifically and efficiently link various technologies and jointly assist modernization.At the same time as the video surveillance deployment is gradually completed at key locations such as classrooms and aisles on campus,the air-conditioning and other IoT devices are also installed in the public area of the campus in the form of a cluster.With the continuous increase of hardware devices,the increasing amount of data,the lack of data sharing and the lack of standard data analysis platforms for management,the current monitoring data collection and analysis cannot be directly and efficiently applied to the automatic control of clustered air-conditioning equipment.Based on the existing campus monitoring data,this paper uses the improved campus monitoring cloud service platform to analyze the number of people in the classroom using the improved human-based detection method.While ensuring that the monitoring data can be efficiently collected and utilized,the analysis result is used as the basis for the automatic control of the air conditioning remote control system,thereby automatically controlling the air conditioning cluster device by using the implemented remote control service.This paper mainly studies and designs from the following three aspects:1.Aiming at the problem of applying the aggregate channel feature to the human body,a multi-channel human body detection method based on Haar-Like prior knowledge is designed because the projection contour of the human body is easily changed and its accuracy is affected.Firstly,according to the projection characteristics of the upper body of the human body,combined with the contour features of the upper body of the human body,it has better robustness,and a human body half-body model is designed.Secondly,a set of Haar-Like feature template families is designed based on the human body half-body model.Combined with the Haar-Like feature template cluster,it is used together to improve the accuracy of human detection.Finally,the Haar-Like features of the design were trained using the Real Adaboost classifier to improve the detection of the detection method.2.In order to solve the problems of monitoring data collection,transmission and analysis faced in the process of intelligent monitoring system on campus,a campus monitoring cloud service platform was designed and implemented.According to the monitoring data processing analysis requirements,the platform is divided into five independent service layers.Through the special design of each service layer,the platform can meet the scalability of services in data collection,transmission and processing,and ensure the low coupling between services.3.Based on the cloud service platform,the device communication processing service and the human body detection number statistics service are designed.Under the support of the data layers of the cloud service platform,a remote automatic control system for air conditioners was realized,and the rationality and effectiveness of the cloud service platform were verified.This paper starts from the computer vision application,and combines it with the current big data environment and"Internet of Everything"to design a complete cloud service platform.Through the analysis and test of the air conditioner remote automatic control system based on cloud service platform,it can be seen that it can control the air conditioning equipment cluster in real time according to the monitoring information,verify the scientific design of the cloud service platform and use the human body detection results for air conditioning.The feasibility of remote automatic control of the equipment,while also improving the utilization of the equipment and reducing the waste of resources.