National Conference on Information and Communication Technologies |
Foundation of Computer Science USA |
NCICT2015 - Number 1 |
September 2015 |
Authors: Uma S, Ganga T |
de27dd79-1d58-4fed-a5b7-34aea22e3383 |
Uma S, Ganga T . A New Iterative Triclass Thresholding for Liver Cancer Image using BFO. National Conference on Information and Communication Technologies. NCICT2015, 1 (September 2015), 5-8.
The idea of this paper is to detect the cancer from the liver image. The shape features of the cancer region are measured and it will be used for further diagnosis. The threshold for image segmentation is obtained by using triclass thresholding method. In this method, based upon the threshold the regions are divided into 3 classes. The first and second classes are foreground and background regions. The third class is a "to-be-determined" (TBD) region. This process is done iteratively and it continued until the preset threshold value is met. To obtain the optimal threshold value this method is combined with bacterial foraging optimization and with variants of bacterial foraging optimization. The result of this method is used for further diagnosis.