No abstract

Computer vision is used in many areas, from stationary video surveillance systems to mobile robots with artificial intelligence. The ability of video sensors to automatically adapt to changes in illumination is significant for high image quality, so the development and research of exposure correction algorithms is an important problem. In the paper an exposure time correction algorithm designed for use in computer vision systems is described. There three main ways to change light sensitivity in video camera is known: controlling the camera's exposure time, changing the sensor sensitivity, and lens aperture size. Each of these approaches affects the properties of the generated images in its own way and has its own advantages and limitations. The most simple to implement approaches are often based on exposure time correction. The authors propose a heuristic algorithm based on image entropy maximization. The results of simulation and testing of the algorithm in real conditions are presented. A comparison has been made with the camera auto exposure mode. Based on the results of the research, it has been concluded that the proposed algorithm has better performance with a similar final exposure time, while it has a low computational complexity.