In order to meet the need of human-machine interaction of the portable equipments

an embedded finger-interaction system is implemented. Its applied algorithms such as skin color segmentation

convex hull computation

fingertip detection and etc have been investigated and the hardware has been designed. Firstly

according to the clustering character of skin color

the human hand is segmented via building the skin color model by analyzing the characters of common color spaces. Secondly

the classic Graham-scan method is improved based on the method of radial-scan

and the convex hull of human hands can be computed quickly. Thirdly

the algorithm of fingertip detection which uses the curve of human finger is discussed. Finally

the DSP and FPGA based hardware structure is introduced. Experimental results indicate that the detecting precision is 95.2% for a natural stretched finger

is 92.6% for the bended finger and is 90.1% under the disturbance of other fingers. The maximum offset of fingertip location is 2.12mm and the location time is 23ms. The system is nature and friendly

has strong stabilization and can satisfy the real-time requirement.