Research And Design Of Human Skeleton Modeling And Gait Detection System Based On Video

Gait is the walking attitude of a person,and is a very important feature to identify people.Gait is the only biological feature that can be perceived from a long distance.Although it is not a significant biological feature,it has obtained a good recognition effect.In recent years,it has attracted much attention in the computer vision field.This paper focuses on the analysis of human motion in video image sequences,discusses human skeleton modeling,gait feature extraction,and classification and recognition.This paper mainly does the following work:1)The research analyzes the current status of moving target detection and gait detection at home and abroad,and conducts research on moving target detection from the aspects of noise reduction,binarization,and burr removal.It points out the most commonly used moving target detection algorithm,namely advantages and disadvantages of background subtraction.Morphological operations are used to solve the problems of burr and hollowness caused by background subtraction.2)By analyzing the video sequence in the human gait cycle,and using Zhang Suen algorithm to refine the human body,the human skeleton is obtained.A human skeleton model based on joint points is established using a model based on human proportion analysis.Using the obtained skeleton model,the position of hip joint,knee joint and ankle joint on the side of human body is obtained by analyzing the location algorithm of lower limb joint.The angle of joint point of lower limb is analyzed.3)Aiming at the limitation of insensitive historical state in the application of HMM,the HSMM was constructed by incorporating historical state information of gait data into the modeling process.The mean and variance are verified by experiments.The results show that HSMM improves the accuracy of gait classification.HSMM model is used to train the parameters of normal gait and abnormal gait.The curve of joint angle of lower limb with time is drawn.Experimental results show that the gait detection algorithm based on HSMM has ideal recognition performance.4)In this paper,a gait detection system based on human gait characteristic data is designed and developed based on Visual Studio development platform based on the idea of module design,and the workflow,structural framework and main functional modules of the system are introduced in detail.