Product definition application using image processing techniques in automation systems

The quality of the final product in any production line for all companies in the manufacturing sector in the world, especially firms producing in the casting industry, is one of the most important factors. In this context, companies are trying to realize the quality control process, which has become the final stage of production, with minimum error. The control towers consisting of sensors are used as the traditional method to carry out the control process. This method which is an electronic component consisting of mechanical sensors can give erroneous results due to some physical reasons and with time. In addition, the use of varying numbers of sensors and different designs for each different product on the production line reduces production flexibility and production speed, while increasing the cost in this process and also reducing the quality in production. For these reasons, traditional method is no longer preferred by companies. In this study, an alternative application which responds to the problems and demands of companies to the traditional method was realized. In the study, an embedded system was created with Raspberry Pi 4B developable card and Raspberry Pi V2 Infrared camera. This system is supported by a physical environment containing a powerful infrared light source with a lens and filter designed specifically to absorb sunlight due to difficulties in processing images of aluminum parts and the sunlight in the environment affecting the image. To program the embedded system, a desktop application has been developed using the C# programming language in a .Net environment. Embedded system was programmed with the Python programming language using the OpenCV library, which contains image processing techniques such as Brightness, Contrast, Hue, Saturation, Thresohld, Erode, Dilate, MatchTemplate and Surf Alghorithm. With the interface implemented, the system can be programmed for 32 different prod-ucts and 32 different areas that can be determined on each product. The connection between the card and the interface has been realized with a Socket applica-tion on the TCP protocol, so the connection has been secured over the Ethernet and Wifi via the wired or wireless meth.