Color image compression/reconstruction by YUV fuzzy wavelets

A color image compression/reconstruction method based on fuzzy wavelets is proposed, where YUV color space is employed. Since the Y space includes more important information than that of U and V space, more efficient image compression/reconstruction can be achieved compared with RGB space based one by employing the condition that the file-size of compressed image of Y space is bigger than that of U and V space. The proposed YUV fuzzy wavelets are based on a partial fuzzification of morphological wavelets, and the operations of the wavelets can be performed by fuzzy relational calculus, partially. To compress/reconstruct the image effectively, a soft threshold scheme can be given as alpha-band, which is an extended concept of, alpha-cut of fuzzy relation. Through an experiment using a color image extracted from SIDBA, it is confirmed that the root mean square error of the proposed YUV fuzzy wavelets is decreased to 78.65% of that of RGB fuzzy wavelets, under the condition that the compression rate is about 0.11.