The Development of a Model of a Number Plate Recognition System Control in Real Time

Abstract

Processing video streams in real-time is one of the most difficult areas of machine intellectualization. Intelligent systems of processing and analyzing video streams are more and more intensively used in various fields of human activity. They are most widely used when dealing with such tasks as securing various facilities, development of traffic management systems, etc. One of the areas of developing systems for processing and analyzing video streams is the development of systems for license plate recognition.Today, the market offers quite a number of plate recognition systems, but there is no system produced in Kazakhstan. After analyzing the known plate recognition systems the following was revealed: firstly, many systems give unsatisfactory results for images of poor quality; secondly, they do not work with complex (diverse) images, since it is difficult to locate the required recognition area in them; and thirdly, they are all aimed at well-defined conditions (lighting, camera angle, brightness, etc.). Their high cost does not allow for mass introduction.The peculiarity of developed system is that depending on the image quality and technical characteristics of the equipment the data processing is performed by automatically selecting a particular model and algorithms which in turn provides the system's rate of

Application of a scenario approach in development of a recognition system of vehicle identification numbers

“…Thus, the choice of this or that algorithm significantly influences the performance of the entire model and the reliability of recognition [1].…”