00000ctm a22000004a 4500 UP-99796217603116339 Buklod 20230503092542.0 m |o d | ta 070717s2000 xx r |||| u|eng d (iLib)UPD-00000179783 DENGII LG 993.5 2000 C65 A48 Amutan, Eunice D. Interaction with onscreen objects using hand gesture recognition Eunice D. Amutan ... [et al.]. 2000. vii, 34 leaves Thesis (B.S. Computer Science)--University of the Philippines Diliman. This paper presents the algorithms used in the design and implementation of a visually based hand gesture recognition system used for the manipulation of onscreen objects. Video images are processed and recognized in real time, enabling the computer to react on gestures. The system can be use to move and rotate objects simply by moving and rotating your hand without wearing any devices or using any markers. The system can be used similar to the mouse, it is not intended to replace it but it is seen as an additional alternative. Learning Vector Quantization (LVQ) algorithms were first used to generate a codebook vector corresponding to the skin and non-skin pixels in the images. The system then use a hand detection algorithm to identify in the image the skin region (clusters of skin pixels) which pertains to the hand of the user. After this, the system use a gesture recognition algorithm to determine the gesture being executed by the user. Each gesture has a corresponding action and these actions are executed upon recognition of the gesture. These algorithms are discussed in detail in the paper. The location of the hand is then mapped to the screen. This could be useful in applications development because the location of the hand with respect to the output screen can be used to manipulate onscreen objects. The system can be used with any background. Experiments show that for the optimal detection and recognition of the hand gesture we must observe the proper camera distance and specific lighting condition. Computer software. Learning Vector Quantization (LVQ) algorithm. Hand detection algorithm. Gesture recognition algorithm. Cenas, Joanne April D. Francisco, Lelani B. Manansala, Marlon S. Santos, Roselyn D. Thesis FI UP UPD DENG-II LG 993.5 2000 C65 A48 Thesis