eng Science and Education Publishing Journal of Computer Sciences and Applications 2328-725X 2018-06-04 6 1 32 37 10.12691/jcsa-6-1-4 JCSA2018614 article Biju Bajracharya bajracharya@bsu.edu 1 David Hua 1 Department of Information Systems and Operations Management, Ball State University, Muncie, USA Image compression methods are used to efficiently reduce the volume of image transmission and storage. Pre-processing of images are done to remove spurious noise or unwanted detail from an image to improve the compression performance. This paper proposes a preprocessing method for image compression based on ±K adjustment to a pixel value that enables high compression ratio without losing visual quality. Visual quality of an image was measured using peak signal to noise ratio (PSNR) as a metric. This method was designed based on mapping table constructed from histogram to identify pixels that hinder high compression ratios. These identified pixels were adjusted by ±k values which yielded higher compression ratios. The designed method had six levels of operations. Higher levels retained most of their original pixel values, thus maintaining higher PSNR values at lower compression ratios. Lower levels achieved higher compression ratios by adjusting more pixels (lower PSNR values). A value of ±1 was used for retaining better original information, while ±2, ±3 and higher were used for higher compression ratios. Preprocessed and non-preprocessed grey scale images were compressed using popular lossless compression algorithms like Deflate, Bzip2, LZWA, and 7zip. Our experimental results show that this method significantly improves compression ratios as compared to compression without preprocessing. http://pubs.sciepub.com/jcsa/6/1/4/jcsa-6-1-4.pdf image preprocessing preprocessor experiment compression digital images image compression PSNR