Abstract | ||
---|---|---|
Block Truncation Coding (BTC) image compression technique is capable of extraordinary computational efficiency for high-framerate configurations under the trend of high image resolution. However, the recent development such as the Dot-Diffused BTC (DDBTC) method still encounters various challenges of visual artifacts, including impulsive noise, false contours and blocking effects. In this study, a new method, the Near-Aperiodic DDBTC (NADDBTC) technique, is proposed to analyze the tradeoff and balance between computational efficiency and visual artifacts. It is mainly because of the quantization levels and the co-optimized Class Matrix (CM) and Diffused Matrix (DM) during image compression procedure. In addition, the tiling method for bitmap generation is optimized for aperiodic compressed results. Experimental results show that this method is capable of excellent image quality and visual perception, as well as processing efficiency, at a level similar to the DDBTC technique by taking advantage of parallel processing in dot diffusion. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1016/j.sigpro.2015.09.030 | Signal Processing |
Keywords | Field | DocType |
Image compression,Block Truncation Coding,Dot diffusion,Digital halftoning,Optimization | Visual artifact,Block Truncation Coding,Control theory,Computer science,Image quality,Artificial intelligence,Computer vision,Color Cell Compression,Algorithm,Bitmap,Quantization (signal processing),Image resolution,Image compression | Journal |
Volume | Issue | ISSN |
120 | C | 0165-1684 |
Citations | PageRank | References |
0 | 0.34 | 9 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yun-Fu Liu | 1 | 277 | 19.65 |
Jing-Ming Guo | 2 | 830 | 77.60 |
Zong-Jhe Wu | 3 | 16 | 1.36 |
Hua Lee | 4 | 109 | 11.38 |