Segmentation of Microcopy Images Using Gradient Path Labeling and Artificial Intelligence Techniques

Recent studies using microbes as model organisms rely on microscope imaging which needs to be complemented with reliable and fast methods of computer assisted image processing. These methods aim at facilitating the extraction of information from images of bacterial populations with single cell resolution, by avoiding manual analysis, which is fastidious, time consuming and subject to observer variances. To isolate single cells in microscopy images, image segmentation techniques are essential. However, segmentation of nontrivial images is one of the most difficult tasks in image processing. In this talk several segmentation methods will be compared and discussed. Artificial intelligence techniques to circumvent the over-segmentation typical of most classical segmentation methods will also be presented. Practical application examples will be shown to illustrate the results of the different techniques.