Font Size: a A A

The Human Geometrical Dimensions Measuring System Based On Ordinary Camera

Posted on:2014-12-18Degree:MasterType:Thesis
Country:ChinaCandidate:Q HanFull Text:PDF
GTID:2298330422968261Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
Since the1950s, more and more people pay attention to traditional Chinesemedicine. China and other countries have begun to conduct research on acupointsvigorously. During the research process, since the location of acupoints are notaccurate enough, we can not study the mechanism of acupoints effectively,researchers begin to suspect the scientific and reliability of acupoints, it’s verynecessary to inspect, supplement and correct the traditional acupoint positioningmethod with the help of actual measurements of human body geometric dimensions.Traditional manual measurement methods are inefficient, can’t get objectivemeasurement results, and also bring a great deal of inconvenience to both testers andperson under test. In response to this situation, we use modern technologies, such ascomputer vision, image processing, develop a convenient human geometricaldimensions measuring system, and achieve the measurements of two-dimensionalfeature sizes by an ordinary camera.The main jobs of this paper are as following:1. This paper designs a hardware platform of the human geometrical dimensionsmeasuring system, using two cameras to collect the human body image and humanhand image, obtaining the distance by an acoustic ranging sensor.2. In order to achieve a high degree of automation, we add gesture controlfunction in this system. A gesture recognition method based on SIFT algorithm isdeveloped. Using skin color detection to select images which are collected by thecamera, using SIFT algorithm to find the sample image which matches the test image,real-time gesture recognition can be realized through above methods, and its output isused to control the human geometrical dimensions measuring system.3. In order to achieve the calibration of the measuring system, using gradientHough transform to extract circles on the planar target. Image processing techniques,such as image segmentation, image annotation, edge detection are used to extractfeature points on the human body; we can get the desired geometry, including height,head width, pupil distance, shoulder width, shoulder height. 4. Designing a man-machine interface by using MATLAB GUIDE. In order toachieve hybrid programming, we compile the system program into dll file which canbe called by other programming languages easily.
Keywords/Search Tags:Human Geometrical Dimensions, Digital image processing, Gesturerecognition, SIFT algorithm, Computer vision
PDF Full Text Request
Related items