Human hand gesture recognition using spatio-temporal volumes for human-computer interaction

With the emerging of the new applications like virtual reality in image processing and machine vision, it is necessary to have more perfect interfaces than mouse and keyboard for human computer interaction. To cope with this problem, variety of tools has been presented to interact with computers. Hand gesture recognition is one of the proper methods for this purpose. This paper presents a new algorithm based on spatio-temporal volumes for hand gesture recognition. In this algorithm, after applying necessary pre-processing on video frames, spatio-temporal volumes are constructed for each gesture. These volumes are then analyzed for matching and feature vectors are extracted in two consecutive stages. Finally a classification algorithm is applied for classification. We examined three different classifiers including k-nearest neighbor, learning vector quantization and back propagation neural networks for recognition. We tested the proposed algorithm with the collected data set and results showed the correct gesture recognition rate of 99.98 percent for noiseless and 92.08 percent for noisy data.