Skeleton Prototype Software Detection and Feature Extraction of Non-intrusive On the gait of Man In Real Time

Hustinawati, Hustinawati and Madenda, Sarifuddin and Kusuma, Tubagus Maulana Skeleton Prototype Software Detection and Feature Extraction of Non-intrusive On the gait of Man In Real Time. Internatioanl Journal of Sport Science and Engineering. ISSN 1750-9823

[img]
Preview
Text
Skeleton Prototype Software Detection and Feature Extraction of_UG.pdf - Submitted Version

Download (2725Kb) | Preview

Abstract

Every human being created by God to the bone structure and shape of the body that affect the attitude of walking human or human gait with a unique karakaterikstik. Human gait analysis can be implemented on a medicine to help doctors diagnose patients with gait disorders, designing a program of rehabilitation and prosthetic design. In the sports field is used for the analysis of human gait trainer consideration in identifying the best techniques to improve the performance of athletes, explaining the ways of safe and effective in performing gait athletes to avoid the risk of injury in athletes, implement the strategy, the pattern of training and therapy to athletes in accordance with physical condition of each. This paper discusses the software prototype skeleton detection and feature extraction of non-intrusive to the human gait in real time. The first stage of human gait analysis is the acquisition of objects of human gait using Logitech webcam in real time VidTmHD 9000. The second stage is composed of pre-segmentation process and the formation of a silhouette. In the pre-process each frame gait resulting from the acquisition of human gait are processed using the method of background subtraction. Subsequent formation of the human body silhouette with a method of filtering, thresholding, dilation, erosion, and inversion. Skeletonizing the third stage, the process of skeleton formation using the method of thinning. The last stage is the feature extraction process parameter measurement of the distance geometry calculations using both feet of space to obtain the coordinates of each frame image of the skeleton of human body parts of both legs at all times. Based on the test results it was concluded the acquisition method of human gait in real time managed to make the acquisition of human gait quickly so that the data of human gait can be directly processed to detect the skeleton and extracting features automatically, quickly and accurately. Background reduction method can effectively separate the object mensegmentasi and ongoing human body from other objects in the image captured by the Webcam. Methods of filtering, thresholding, dilation, erosion, and the inversion result of the human body silhouette running. Method of skeletonizing and thinning of the human body silhouette running to produce the skeleton of the human body is almost like the real form. Calculation of the features of human gait is the distance between the front legs to the back of the heel, the difference in height between the front foot with the heel of the hind legs, the distance between the front knee to knee and back foot difference in height between the front knee behind the knee can be performed automatically without operator intervention or non-intrusive in a relatively fast

Item Type: Article
Uncontrolled Keywords: Gait; Realtime; Acquisition; Skeleton; extraction;
Subjects: A General Works > AI Indexes (General)
Divisions: Fakultas Teknologi Industri > Program Studi Teknik Informatika
Depositing User: Mr Reza Chandra
Date Deposited: 28 Feb 2014 12:04
Last Modified: 28 Feb 2014 12:04
URI: http://repository.gunadarma.ac.id/id/eprint/1283

Actions (login required)

View Item View Item