Cloud-Based Information System for Evaluating Caverns in the Process of Blasting Metal Surfaces of Details

Various methods for the surface treatment of metal parts have been proposed. However, they do not objectively assess the treatment quality in real-time or employ computer methods for processes such as monitoring and quality control. This study aims to develop a method for determining the average size of elements on a digital image and to implement a cloud-based information system for evaluating caverns in the surface blasting process of metal parts. Based on the digital image of the material surface, the leading brightness-control frequency is estimated. During the research, a mathematical apparatus is developed based on differential calculus and employed to determine the leading frequency. This method differed from the existing methods, as the local value of the phase of the analyzed oscillations was independent. Based on the proposed method, a computational algorithm for determining the average grain size on a synthesized image was developed, the quality of which was improved by implementing iterative refinements. The effectiveness of the proposed algorithm was evaluated on real macro photographs of the treated metal surface (i.e., a shot-blasted brass plate). The experimental data confirmed the achievement of acceptable parameters for determining the average size of elements on a digital image. Furthermore, the suitability of the synthesized algorithm for software implementation in the form of a Platform as a Service was verified. The proposed cloud information system, as a computer vision tool, estimated the size of caverns on the digital image of the processed metal surface. Further, the zones of the received frame were marked according to the average size of the caverns. Additionally, the system allowed the automation of the cavity evaluation process during the shot blasting of the metal surfaces. The proposed method will facilitate the automatic evaluation of the graininess of metal surfaces using specified functions with a proposed computer vision tool.