ABSTRACT
V/M Graph is a logical method based on our thought that could describe all the information and the relation between all the variable using a graphic structural in processing the picture with the extracted background for furthermore processing in the identification system. This V/M Graph method can be use to detect a moving object in the video from a camera. V/M Graph is very useful to analyze a video or a picture so that it could detected a motion on human body and identify it to be recognize. V/M Graph technology can be applied into a system which can help us to identify how someone to walk without have to see it physically.
I have made a detection system which could indetify on how someone to walk without have to see it physically on this Final Project. If my detection system cannot identfy it, it will give an information that individual is not recognize. My purpose on creating this system is to detect someone through their walking gesture without have to see it physically, eventhought they physically changed my system could still recognize it. Clasification methode to identifiy a walking pattern using V/M Graph.
From the analyzed and testing, the system is able to achieve an average accuracy rate of 86.67 % with a test system to changes in light intensity. When testing the system with optimal threshold accuracy results with the threshold of 90 % from 1000 to 1300 pixels. And when testing the system with a wide threshold value label pixels 0,25 largest system gives an accuracy of 90 %. Target system performance has been achieved is a system can recognize and identify gait with optimal accuracy rate.
Keyword: V/M Graph, walking pattern, human identification.