EE368 Project FACE DETECTION AND GENDER RECOGNITION

The first step in performing colour-based segmentation is choosing an appropriate colour space in which to operate from the wide variety of choices such as RGB, HSV, CMYK, YCbCr, etc [1]. Of these, RGB (red-greenblue) and HSV (hue-saturation-value) have been the most widely used. Figure 1 illustrates the geometries of the two spaces. By way of example, HSV representation has certain advantages over RGB when it comes to face detection. As Garcia et al [2] note, skin colours are sensitive to the lighting condition. In the RGB space, each of the three components may exhibit substantial variation under different lighting environments. In HSV space, however, the hue and saturation components are virtually unchanged. Figure 3 shows the histograms of RGB component values of both face and non-face pixels over all seven training input images. Similarly, Figure 4 shows the histograms of the same images in HSV space, where the S component and in particular the H component are well-clustered for face-pixels, while H and S are spread over a wide range for the remainder of the image. This observation favours using an HSV colour space if only a simple thresholding colour segmentation is desired. G