Now days, the development of ontological models for creating edges and their contours for moving objects in real time or close to it is an urgent task. An ontological model for implementing this process is shown in the article. The main algorithms for detecting object edges and constructing contours in an image and program codes for their implementation are considered in the article. It is noted that the Canny algorithm is the best for recognizing edges. At the same time, its serious drawback is determined, which consists in the fact that with insignificant movement of objects, more than 50% of information about the contours is lost.