Compression of Cartoon Images
A novel technique specific for the compression of low color palette images
- 84 pages
- 3 hours of reading
The book introduces an innovative technique for compressing cartoon images, leveraging distinct color regions through a spatially-based algorithm. It combines Binary Run-Length Encoding with Huffman Encoding for efficient storage. This lossy compression method eliminates artifacts and antialiasing, while a post-decoding edge-restricted blur filter smooths edges. Additionally, it details algorithms for gradient detection and their applications, achieving significant file size reductions—averaging 13.75 times smaller than PNGs and 7.45 times smaller than JPEGs, with an impressive bit per pixel ratio.