Quality control is an important but time-consuming process. In the age of Industry 4.0, where we strive for the highest possible level of automation, it is important that the quality of the assembly can be monitored in the absence of humans. To this end, we need a measurement system that allows us to check the quality and suitability of the assembly of the Raspberry Pi product, which is assembled by a robot in the production line. Therefore, as part of this task, we have developed a machine vision system that allows the detection of individual components and subassemblies of the selected Raspberry Pi product. In particular, we focused on the component and subassembly recognition algorithm based on the principle of component color features and compilation of images of the reference product. Reference images of subassemblies and assemblies are used as a template for comparison with real-time images of assembly operations of an actual product and its subassemblies. The image algorithm analyzes, evaluates and provides the result of the adequacy of the assembly. A prototype of the machine vision system was made, and the main parameters and settings of the camera, lighting, and measuring point are given. In the experimental part we analyzed and tested the measurement methods and the algorithm.
|
