Image processing using the HSI color space

Traditional interfield spatial filtering in the YIQ color space is compared with new spatial filtering techniques developed using the HSI color space. Results based on an extended version of the CCIR impairment scale are presented. The present experiments are based on renderings of computation grids obtained from the MSU/NFS Engineering Research Center for Computational Field Simulation. It is shown that there are three significant points to be considered when performing spatial filtering in the HSI color space for the purpose of flicker reduction. First, sharpness degradations are normally considered more objectionable than flicker degradations. Second, luminance and intensity filtering are visually equivalent. Third, saturation processing can be used to introduce appropriate colors into the gray scale fringe regions that are generated by intensity processing, but hue processing must be used in conjunction with this to avoid the red fringe effect. Therefore, the marginal improvement in image quality does not normally justify the additional computational effort required