Least-Squares Halftoning Algorithm Based on Image Segmentation

Abstract

This work presents a method based on the image content for digital halftoning using K-means clustering theory. Our algorithm applies to both a printer model and a model for the human visual system (HVS). The method strives to minimize the perceived error between the continuous original image and the halftone image. First, the gray image is partitioned into two, three and four regions using K-means image segmentation method, whose performance depends on the selection of distance metrics. Next, the statistics of average gray value of each clustering is calculated. Each clustering uses the least-squares model-based(Lsmb) algorithm to obtain halftone image. Finally, analysis and simulation results show that the proposed algorithm produces better gray-scale halftone image quality when we increase the number of clustering with a certain range. A performance measure for halftone images is used to evaluate our algorithm. The value of MSEv, WSNR and PSNR for two partitions is almost the same as that of the Lsmb algorithm, but for three and four partitions that the proposed algorithm achieves consistently better values of MSEv, WSNR and PSNR than the Lsmb algorithm.