Hand Gesture Recognition Using Statistical Analysis of Curvelet Coefficients

Hand gesture tracking is very challenging task with different backgrounds and illumination conditions. However, the challenge becomes more tough when the hand gesture is moving in all 3600 direction. In the earlier work as studied in literature survey, it has been found that the location, rotation and size invariance has been the major problem to deal with. In the proposed work, efforts are made to overcome the size, rotation and location invariance using statistical tools. A discontinuity point affects all the Fourier coefficients in the frequency domain. The presented algorithm is based on statistical analysis of curvelet coefficients, and we take the minimum standard deviation, minimum class variance, minimum differential entropy and maximum PSNR.