Computer vision‐based automated defect detection in ceramic bricks

Nowadays, the development of cost‐effective, data‐driven technological processes using telecommunication technologies is essential. One of the focuses is on automating the process of evaluating the manufactured goods' quality. Vision‐based technology is now becoming increasingly used for monitoring purposes. Despite its advancements, computer vision technology has practical limitations. These include the physical characteristics of the measuring process, features specific to the technological procedures, and constraints related to software and mathematical algorithms. Among the cutting‐edge approaches, optical methods combined with neural network algorithms (NN) stand out. This significance is particularly evident because numerous industries continue to depend on manual defect identification methods, which are labour intensive, slow, and subject to human subjectivity. The article introduces a novel approach based on computer vision methods. It outlines an automated optical inspection system designed to detect defects in bricks on a transport belt during the production process. The article presents the processing algorithms used and discusses the results obtained.