STUDY OF THE IMPLEMENTATION OF AUTOMATIC SYSTEMS WITH COMPUTER VISION TECHNOLOGY IN MANUFACTURING

Abstract

An overview of the use of computer vision systems, the complexity and problems of their development and use in modern industries that require automation of labor, and the introduction of the latest technologies to control the technological process. The main promising areas for the implementation of computer vision systems are highlighted. The main reasons for the complexity of building such systems, which have both the physical nature and the complex nature of computer vision as a discipline, are identified. The interaction of such systems with programmable controllers is considered. The use of programming languages for both controller programming and computer vision development is examined. These are C++ and Python programming languages for software development that have corresponding libraries that implement recognition and learning of systems by implementing machine learning, as well as ST, IL, LD, FBD, and SFC languages for controlling programmable controllers. Likewise, the implementation of basic requirements for computer vision systems, such as flexibility and scalability, is possible owing to C++ and Python languages with corresponding third-party libraries. The programming language in this case becomes a convenient tool for actualizing pre-implemented complex functions. Attention is drawn to the fact that automatic systems using computer vision are discrete-analog and require appropriate technical devices and programmable controllers. The technical elements that they should consist of are also considered.