In many vision problems, the performance of the seg- mentation step is highly dependent on the algorithm selec- tion and its parametrization. These tasks are tricky and time-consumming. In this paper, we present an approach to perform task-oriented segmentation based on segmenta- tion algorithm parameter tuning and learning techniques. We propose a scheme that, for each segmentation algo- rithm to test, first extracts optimal parameters and second learns the region labeling according to the segmentation task. This supervised approach uses two kinds of ground- truth data: manual region-based segmentations and seman- tic region labels. The first step consists in extracting optimal segmentation algorithm parameters by using a closed-loop optimization procedure, an evaluation metric and ground- truth (manual region segmentations). During the second step, region classifiers are trained based on ground-truth annotations (semantic region labels) to allow segmentation labeling. This knowledge (i.e. optimal parameters and learned region classifiers) is then used to produce an op- timized class-based segmentation of new images. Our main contribution is to propose a methodology to easily set up the segmentation task in vision systems. Our method only requires the user to provide segmentation algorithms and labeled ground-truths. The experiment on a biological ap- plication shows that our optimization scheme is reliable for different state-of-the-art segmentation algorithms. Evalua- tion of results also demonstrates that the optimized class- based segmentation achieves a better level of accuracy than the non-optimized class-based segmentation approach.