Hand Morphological Feeature Analysis Based On Image And Mark Watershed Algorithm

Human dimension data is an important basic data resource.As an important part of the body,hands are responsible for most of human actions.Real and effective hand data is of great significance for hand feature analysis,related product design,medical research,and criminal investigation.In order to solve the problems of delayed hand measurement data,low efficiency of manual measurement and high cost of 3d measurement,a method of hand size acquisition based on image and mark watershed algorithm is proposed in this paper.Then the accuracy of the obtained data is verified,the hand shape and characteristics are analyzed.This method has the advantages of low cost,fast and accurate measurement,and low requirements for external environment and testers.The main research contents of this paper are as follows:Taking young male hands in Zhejiang as the research object,50 subjects are selected,and the left and right hand size data are collected by manual measurement.Through the correlation analysis of the left and right hand data,it is determined that the right hand is the image acquisition hand.Aiming at the possible influencing factors in hand image acquisition:lighting,distance,stability,etc.,a special hand image acquisition box is designed to collect the right-hand images of 200 subjects.It not only ensures the efficiency and quality of image acquisition,but also fully considers the portability and manufacturing cost.Aiming at the texture characteristics of the hand and comparing the results of the traditional image segmentation algorithm,a method of extracting feature points of the hand using the mark watershed algorithm is proposed This method not only has a good response ability to the weak edge such as the shadow area of the hand contour and knuckle marks,but also can avoid the over-segmentation problem caused by the complexity of the hand texture,obtain the continuous curve of the hand contour and knuckle marks.The steps to obtain the true distance of 24 hand measurements based on the image and the marker watershed algorithm are as follows:firstly obtain the contour curve of the outer hand,extract the key points of the root,fingertip,wrist and knuckle position.Then calculate the pixel distance of each hand size according to the extracted hand feature points and measurement item definitions,convert the real distance according to the camera calibration distance.For the extraction of finger-root and fingertip key points,a method to obtain the peak curvature point of hand contour is proposed.The method of extracting key points of knuckle position is to perform a linear fit on the knuckle curve coordinate data,find the intersection point of the fitted straight line and the contour of the finger,so as to determine the key points of knuckle position.Finally,20 subjects' hand images are randomly selected from 50 manually measured subjects,and the data obtained by the manual measurement method and the image-based and marking watershed algorithm are compared.The result shows that the method designed in this paper could quickly and accurately calculate the hand size of 24 items.The morphologic characteristics of 24 right hand data obtained from 200 young men in zhejiang are analyzed.Four factors with the greatest influence on hand shape are determined,and hand shape is divided into five categories by GS algorithm and k-means clustering algorithm.Then explore the relationship between the commonly used hand sizes and the ratio of hand length to height.The linear relationship between them is clarified.Finally,in order to reduce the requirements on users and improve the degree of automation and integration of hand size measurement,a complete hand morphology and feature analysis system is developed to integrate all the research results in this paper.Through this system,a series of operations such as hand image acquisition,automatic hand size extraction,data storage,hand shape and feature analysis can be conveniently realized,which further improves the efficiency of hand size measurement.This paper's research on hand data collection and hand morphological character analysis updates the hand data of young men in Zhejiang,and provides new ideas for large-scale data collection.It is of great reference significance for consumers to buy hand products such as gloves,hand tools and manufacturers to make these goods.At the same time,the regression relationship between height and hand length of young men in zhejiang has theoretical and practical significance in the field of criminal investigation.