Detective product classification using image processing is a powerful application of computer vision to identify, classify, and organize products based on their visual characteristics. In this study, the system for detecting and classifying defective products was developed by integrating into the production system with a camera and a UR5 collaborative robot. The inspection and detection of defective products is performed on the two bottom surfaces of cylindrical workpieces by recognizing defective images appearing within the contour of the workpiece surface using an image processing program programmed in Python with the help of the OpenCV library. The operation of the subsystems is logically controlled by transmitting/receiving data through establishing a communication connection. The results of the application of the system to the problem of identifying and classifying defective products that can appear on any two surfaces of a cylindrical workpiece have been evaluated and have a high success rate of classification.