Artwork Defect Detection Based on Computer Vision

Abstract:

Art identification becomes more and more sophisticated today, so the conventional testing or visual inspection is difficult to ensure accurate detection of small defects. This paper presents a computer-based defect detection method for artworks. It adopts small components removing method to remove image noise and any parts interfering measurements, and maximum variance method as threshold method of the detection system. Pixel and sub-pixel level edge detection methods are used to overcome the shortcomings of traditional methods to complete the defect detection. Experiments show that the scheme is feasible, and has high accuracy and shorter detection time.

• Periodical: Advanced Materials Research (Volumes 846-847)
• Pages: 1234-1238
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.846-847.1234
• Online since: November 2013
• Authors: Qiu Shi Li*
• Keywords: Artwork, Defect Detection, Noise Removing, Threshold Segmentation

References

[1] Sun Yinan, Liu Weijun, Wang Yuechao, etc. An improved Hough transform method for circle detection [J]. Computer Engineering and Application, 2003, 20: 35~37.

[2] LEAVERS V F. The dynamic generalized Hough transform: Its Relationship to the probabilistic Hough transform and an application to the concurrent detection of circles and ellipse[J]. CVGIP: Image Underst, 1992, 56(3): 381~398.

DOI: 10.1016/1049-9660(92)90049-9

[3] LIU Jianqun, Li Shiyong, Kuang Hui, etc. The application of Improved Randomized Hough transform in detection of multi-circle [J]. Microcomputer Information. 2007, 10(1): 288~290.

[4] Li Yan, Peng Jiaxiong. Sub-pixel edge detection based on spline interpolation and D2 LOG operator [J]. Huazhong University of Technology. 2000, 28(3): 77~79.

[5] Tabatabai A. J, Mitchell O.R. Edge location to subpixel values in digital imagery[J]. IEEE Trans On PAMI, 1984, 6(2): 188~201.

DOI: 10.1109/tpami.1984.4767502