A Intelligent Computer Input System Based On The Hand Actions

In the 21st century, the quality of human livings has been improved greatly by the development of the science and technology. In the research study of image processing, motion observation and intention understanding attract great attention. In theory, the method of motion recognition must be efficient and precise and then the system can be developed for data processing and intelligent applications.By now, the influencing factors of motion recognition include the diversity of motions, complex backgrounds and target occlusion. The generality and universality of detection algorithms designated for solving these problems are not promoted for various cases.The gesture detection is designated as the research breakthrough in this paper. We build a model for gesture detection based on the recognition mechanism of human behaviors. The detection procedure is composed of contour detection based on the skin color, contour noise filtering and smoothing, corner detection, contour dithering filtering as well as action tracking. And the intention understanding of human gesture is discussed and studied based on the recognition model.An intelligent online system of gesture prediction and intention understanding based on the cognitive model is presented for characters input. This system is efficient and precise for entering the characters at the condition of low-vibrating and complicated gestures. The experimental result demonstrates the validity and practicality of our proposed system.