Facial Feature Analysis System Based On Huawei HiSilicon 3536 Development Board

The rapidly evolving Embedded System lays the hardware foundation for face detection and facial feature analysis in complex environments.Based on the Huawei HiSilicon 3536 development board,this paper builds a facial feature analysis system,and based on the hardware platform,the algorithm based on LAB and MLP is selected for face detection and random forest cascade regression algorithm to train the facial feature point model.The eye features were analyzed,and finally the algorithm was implemented on the hardware platform and tested.The main work of this paper is as follows:1.The background and significance of this paper are studied.The facial feature analysis algorithm and the research status of embedded system at home and abroad are analyzed.2.Learn more about the functional structure of the Huawei HiSilicon 3536 development board and the performance of the HIKVISION infrared camera.The embedded operating system is transplanted for the hardware platform,and the cross-compilation tool is installed,and the programming environment under the Windows system is configured.3.The face detection based on the combination of LAB and MLP is used to save the detection time by using the coarse to fine detection method.The face feature point model was trained by random forest cascade regression algorithm,and the closed state of the eye was analyzed by threshold judgment.In the HELEN face database,a comparison experiment with various algorithms is carried out.The experimental results show that the algorithm used has high positioning accuracy,which meets the real-time requirements of the system.4.The overall design of the system was completed,and the algorithm-related libraries(OpenCV and Dlib)were cross-compiled and transplanted,and the speed performance of the system was tested.The test results show that the system can accurately and accurately in the case of day and night illumination changes.Displaying the closed state of the eye satisfies the requirements of real-time and accuracy of the facial feature analysis system.Based on the Huawei HiSilicon 3536 development board and HIKVISION infrared camera,this paper designs a facial feature analysis system to reduce the impact of complex detection environments such as day and night illumination changes.The detection speed and accuracy are improved.