Gesture recognition television control system using color histograms

Due to gesture recognition, smart TVs that are controlled by simple motions have been steadily developed and improved. Important problems, however, remain unresolved. In comparison to low-price remote controllers, it is difficult to achieve competitive prices and performances for diverse and complex gesture-recognition systems with the use of costly 3D motion sensors. This study therefore aims to implement a high recognition rate by using only one camera and color information instead of an expensive 3D motion sensor; additionally, the high recognition rate achieved with this technique will make TV control more convenient than a remote controller. We used the Levenshtein algorithm to recognize gesture motion, we used information from the color histograms to track an object, and the HSV color model was used to decouple chromatic information from the shading effect. This study employed an FPGA implemented circuit designed by Verilog HDL.