Equalized CR-dropping for Image Compression
Chun-Hsiu Yeh, Jau-Ji Shen, Jinn‐Ke Jan · 2015
This paper proposes a new idea for image compression that neither the additional information nor the complex formulas for image decompression are needed. Our method includes removing image pixel values from the equalized selected rows and columns for image compression and uses known adjacent pixel values to calculate unknown pixel values for image decompression. We experimentally found that the image quality after compression and decompression by our method is better than that applying joint photographic experts group (JPEG) method for image compression and decompression. Furthermore, the quality of the decompressed images can be higher than that of JPEG images even when the compression rate is higher. Considering the LENA image, when the compression rate is 25%, the image quality in our method is 33.45 dB and that of JPEG images is 32.83 dB, when the compression ratio is 6.25%, the image quality in our method is 27.93 dB and that of JPEG images is 27.49 dB.