Real-Time Automatic Video Inspection System for Piece Products Marking

The paper considers the problem of real time automated video inspection of piece products. The architecture of automatic video inspection system is proposed, the temporal characteristics of its action are investigated. The possibility of analyzing images of objects after transformations in grayscale and Lab space, the use of Hough transforms and second-order moments to frame object localization is considered. The possibility of using machine learning algorithms for real time automatic video inspection of industrial products was studied.