FORI-CSDR - A New Approach for Context Sensitive Image Data Reduction
This paper presents the theory of the FORI-CSDR algorithm (Focus On Regions of Interest — Context-Sensitive Data Reduction). The FORI-CSDR algorithm is able to reduce the amount of image data in dependence of the image context. Within the important image regions nearly no image data will be reduced. The image data of the unimportant image regions are reduced in a very strong way. The data reduction factors for the important and the unimportant image regions can be chosen independently. The FORI-CSDR (focus on regions of interest — context sensitive data reduction) algorithm is able to increase the total data compression rate by varying the image quality settings block-wise, depending on a determined DOI (degree of interest) value.
KeywordsVariable Length Code Quantization Unit Pixel Block Quantization Table Object Edge
Unable to display preview. Download preview PDF.
- 1.Crouse, M and Ramchandran, K, Joint thresholding and quantizer selection for transform image coding: entropy-constrained analysis and applications to baseline JPEG, Proceedings of the society of photo-optical instrumentation engineers (SPIE) VOL. 2847 (1996), 356–364Google Scholar
- 2.Mitchell J L and Pennebaker, W B, MPEG Video: Compression Standard (Digital Multimedia Standard Series), in Chapman & Hall USA (1996)Google Scholar
- 3.Ortega, A and Vetterli, M, Adaptive quantization without side information, Int'l. Conf. on Image Proc. ICIP '94; Austin / Texas (Oct. 1994)Google Scholar
- 4.Ortega, A and Vetterli, M, Adaptive scalar quantization without side information, IEEE Transactions on Image Processing (1997)Google Scholar
- 5.Strohbeck, U and Jäger, U and Macgregor, A E, Context-sensitive image data reduction by FORI, Proceedings IWK '98, 43rd International Scientific Colloquium; Technical University of Ilmenau / Germany (1998), 372–377Google Scholar