This Report presents a state-of-the-art implementation of lossless image compression algorithm LOCO-R, which is based on the LOCO-I (low complexity lossless compression for images) algorithm developed by weinberger, Seroussi and Sapiro, with modifications and betterment, the algorithm reduces obviously the implementation complexity. Experiments illustrate that this algorithm is better than Rice Compression typically by around 15 percent.
The LOCO-R algorithm is based on predictive compression. During compression, the pixels of the image are processed in raster scan order. Specifically, y is incremented through the range[O,ht-l], and for each y value, x is incremented through the range[O,wd-l]. (Thus, the y dimensions the slowly varying dimension.)The first two pixels, with coordinates (0,0) and (1,0), are simply put into the output bit stream encoded. For all other pixels of the image, the processing that occurs can be conceptually divided into four steps:
- Classify the pixel into one of several contexts according to the values of (usually 5) previously encoded pixels.
- Estimate the pixel value from (usually 3) previously encoded pixels, and add a correction( called the bias), which depends on the context.
- Map the difference between the estimate and the actual pixel value to a non-negative integer, and encode this integer using Golomb's variable length codes.
- Update the statistics for the context based on the new pixel value.
HARDWARE AND SOFTWARE REQUIREMENT:
Operating system : Microsoft Windows 98, ME, 2000 Professional, 2000 Server, XP, 2003 Server, or Vista
CPU: Intel Pentium IV 1.2GHz. 2GHz or more is very recommended for desktop PC. Intel Centrino 1.5GHz or more for notebook.
RAM : 256MB. 512MB or more is very recommended. Increase your Windows swap file size if necessary.
Display: A 24-bit color monitor with 1024x768 resolution is recommended.
Hardware accelerator is not required.
- For applications/system where storage space is an important issue.
- Quick loading and processing of images.
- Sending and receiving of images over an network in lesser time.
- For image hosting systems such as Picasa or Flickr where a large amount of images are uploaded daily.