Study Of Vision-based Gesture Recognition Algorithm

The use of hand gestures has become an important part of human computer interaction (HCl) in recent years. Hand gestures play a natural and intuitive communication mode for all human dialogs. The ability for computer to visually recognize hand gestures is essential for future human computer interaction. However, vision-based recognition of hand gestures is an extremely challenging inter-disciplinary project due to following three reasons: firstly, hand gestures are rich in diversities, multi-meanings, and space-time varieties; secondly, human hands are complex non-rigid objects; thirdly, computer vision itself is an ill-posed problem. It involves several fields such as Pattern Recognition, Image Process, Psychology and Cognitive Science. It is also tied up with Human-Computer apperceive interaction. This paper presents a vision-based hand gestures recognition algorithm from points of pre-processing, feature extraction and recognition of hand gestures image.In the part of pre-processing, the paper firstly performs image smoothing, then performs image sharpening in the means of Laplacian operator, and last we get the binary image of the hand gesture.In the part of feature extraction and recognition, we get the connected gesture contour in the way of edge detection based on 8-connected boundary tracking, then the feature extraction and recognition are performed based on two methods: Shape Feature and Fourier Descriptors. As for the feature extraction and recognition based on Shape Feature, gross classification is carried out according to the direction and number of the finger in the gesture. Sequentially, Template Matching is applied to implement the Fine Classification based on the Similarity after the shape feature vectors of the gesture are extracted from the boundary image and binary image. As for the feature extraction and recognition based on Fourier Descriptors, the paper presents and analyzes Fourier Descriptors which do not change with translation, rotation and scale change, and which are also independent of the location of the beginning point of the image edge. The Fourier Descriptors and Euclidean Distance are applied to the recognition ofthe alphabet gesture. We can judge the resemblance between input image with each image in database by computing the distance between their feature vectors, and we classify the image into the sample class that has the shortest distance.The experiment result proves that the gross classification based on shape feature can remove the improper gestures, reduce the matching time and improve the recognition rate. And the recognition ratio is proved 89.6% accuracy in case of the translation, rotation and scale change owing to the application of the Fourier Descriptors.