Design of a Neural Network Classifier for Separation of Images with One Chromosome from Images with Several Chromosomes

In automatic processing of chromosomes, when chromosomes are segmented from image as objects, some of chromosomes that are close to each other (touching chromosomes) or overlapped, are detected as one object. Objects that have more than one chromosome, must be processed separately from those have one chromosome. One stage of automatic processing of chromosomes is automatic separation of multi-chromosome images from one-chromosome images. In this research, a solution is found using feedforward neural networks. Generally, obtained results are satisfactory and better results can be obtained by more processing.