Determining the Best Colour Axis for Detecting and Extracting Segments in Vision Robotics

International audienceOur laboratory develops a system of assistance with a mobile robot to help disabled people in their current life. The environment is characterized by structured elements like segments. In order to permit the localization and the navigation of a mobile robot within an interior environment, we have built a stereoscopic sensor and implemented all the algorithms which allow to obtain 3D coordinates of real objects from data images. Processing on the images leads us to have segments on the two images. These segments are obtained from two images issued of a color stereoscopic sensor put on a mobile robot The matching technique uses two classifiers. We calculate two probabilities to decide if this pairing is suitable. One probability is obtained with a neural classifier; the other is given by a bayesian method. To compute the probability, the two classifiers use the difference between lots of features calculated on the two segments. A previous work, performing on grey level images, has shown the interest of this approach and led us to select eight parameters. Now we work with color cameras, so we have to use this information to increase the good matching percentage. In this paper, we present two methods in order to determine the best color axis. The first method is based on detected segments. In the second method, we extract a sample. By the use of the sample image we will perform the computation of color features to increase the segmentation