A Fuzzy Clustering Approach for Supervision of Biological Processes by Image Processing

We present in this paper a method of image segmentation (T-CAAR) based on a fuzzy logic tree. The construction of the classes corresponding to regions of interest has been made by an upset of membership function and a fusion operator into a tree structure. The comparison of this method with fuzzy c-mean, mean shift and watershed allows using an appropriate technique for biotechnological bioprocess concerning the biomass production. The automatized CAAR method incorporates spatial information which yield the result more accurate and more robust to noise. The data structure allows reducing the CPU time and affecting heavily the result.