Celik, AgitUgurlu, YagizGozukucuk, EmreSonmez, Elena Battini2024-07-182024-07-182019978-1-7281-3964-7https://doi.org/10.1109/ubmk.2019.8907191https://hdl.handle.net/11411/78024th International Conference on Computer Science and Engineering (UBMK) -- SEP 11-15, 2019 -- Samsun, TURKEYThis paper tackles the non-conventional problem of industrial inspection for sheepskin quality checking, which is a sub-field of the leather garment industry. Image processing methods are used in the production lines of several industrial fields for quality checking and production. However, the number of automatic systems is still very limited due to the complexity of the problem, which requires to be addressed. This paper proposes a new neural network-based algorithm to perform the automatic detection of defective parts on sheepskins. The presented method is tested on a newly created database of sheepskins.eninfo:eu-repo/semantics/closedAccessAutomatic SystemQuality ControlSheepskinMask K-Cnnİmage Segmentationİnstance SegmentationConference Object2-s2.0-8507623429310.1109/ubmk.2019.8907191402N/A398N/AWOS:000609879900075