Identifying Hand Gesture Images by Using Genetic Algorithms

A genetic algorithm (GA) is an optimization algorithm that simulates the hereditary phenomenon of natural life. Although GA has been applied to image processing, it has not been studied to apply for matching hand gesture images. In this paper, a method to use GA for the matching process of hand gesture images expressing numbers is proposed. The processing algorithm including GA for hand gesture images is evaluated experimentally by changing the parameters of GA. It is shown that each person can be distinguished from other persons by the fitness value of GA. Hand gesture images to express each number are used for evaluation. Next, the personal identification to search the most suitable hand gesture image in the database of hand gesture images by using GA is proposed and evaluated experimentally. As a result, the effectiveness of the proposed method using GA for matching hand gesture images and identifying each person is confirmed.

• Type: Conference
• Year: 2007
• Where: MVA
• Authors: Tetsuji Kobayashi, Norifumi Machida
• Keywords: Computer Vision, Hand Gesture Images, Images, MVA 2007, Suitable Hand Gesture