An automatic tissue characterization system is always in great demands by pathologists. However, the existing systems are either too time consuming or impractical to analyze liver tissue images. In this paper, we have developed a highly parallel and effective system based on color image segmentation to analyze liver tissue images. To simplify the complex tissue segmentation problem, we first utilize the achromatic information (the intensity) to coarsely segment tissue image, then make use of the chromatic information to classify segmented tissues into different tissue classes. Thus, this automatic system includes an unsupervised probabilistic relaxation classification process and a supervised Bayes classification process. Because the invariant properties of liver tissue image are fully utilized, the computational complexity is greatly reduced. Experimental studies also show reliable results in liver tissue classification for clinical applications. Index Terms, Color image, segmentation, classification, probabilistic relaxation, Bayes classification, voting logic, rank filter.
|Number of pages||9|
|Journal||Biomedical Engineering - Applications, Basis and Communications|
|Publication status||Published - 1993 Jan 1|
All Science Journal Classification (ASJC) codes
- Biomedical Engineering