Abstract: The K-means clustering of pixels intensity to segment the PCB image to a binary form, hierarchical clustering and separation of elements on the PCB image, flood-filling of chains for their comparison with the reference ones, three methods of determination of the turn angle for alignment of the board, subtraction formulas, algorithms to rotate the image to its normal position, algorithms to build the image of the distributed cumulative histogram are considered in this paper.
Roman Melnyk, Roman Kvit, "Defects Detection in PCB Images by Clustering, Rotation and Distributed Cumulative Histogram," WSEAS Transactions on Circuits and Systems, vol. 22, pp. 86-97, 2023, DOI:10.37394/23201.2023.22.11
Roman Melnyk, Roman Kvit. Defects Detection in PCB Images by Clustering, Rotation and Distributed Cumulative Histogram.
WSEAS Transactions on Circuits and Systems. 2023;22:86-97. 10.37394/23201.2023.22.11