Identification of Hand Region Based on YCgCr Color Representation

摘要

For hand gesture recognition, the quality of gesture region segmentation has a direct impact on the recognition and classification results. This paper introduces a color balance method to correct the image color bias of the input images in the RGB color space, and to remove the influence of the correlation of the luminance information contained in the RGB space. We process the input images in the YCgCr color space so that brightness and chrominance are separated. Efficient processing techniques are then proposed in order to remove noisy and skin-like regions in the hand skin segmentation results. Experimental results show that after color balancing and converting to YCgCr color space, it is more effective to detect skin regions from the color images with complex backgrounds, also has more robust illumination adaption.