Perceived Image Gamut Based on Human Visual System and Wavelet Transform

摘要

As some images perhaps contain pixels whose individual colors will not be perceived if they are below the spatial threshold of human visual system (HVS), the paper presents a wavelet transform method for color image considering HVS characteristics to reduce the redundant color information and then construct perceived gamut of image (PGI), which will be used for effective image-dependent gamut mapping. Taking into account the Contrast Sensitivity Function (CFS) of HVS model, the color information reduction is implemented in some subbands of the wavelet transform domain. Then the image gamut was visualized using the general convex hull algorithm and the constructed PGI will only indicate the perceived part of color image. We preliminarily establish a method for reliably determining PGI, which is help to developing robust and psycho-visually justified image gamut mapping algorithm in the future.