Industrial Weak Scratches Inspection Based on Multi-Feature Fusion Network
For scratch detect, our contribution contains:
- multi-feature fusion: dual-attention mechanism and context fusion
- auxiliary loss: enrich the context information and accelerating the training
- real-world industrial datasets
Original defective image image histogram-based thresholding[1]
Moment-preserving thresholding[2] Kittler[3]
ISODATA[4] Yen[5]
Garbor Based[6] Our Method
-
histogram-based thresholding 谷底最小值[1]：C. A. Glasbey, "An analysis of histogram-based thresholding algorithms," CVGIP: Graphical Models and Image Processing, vol. 55, pp. 532-537, 1993.
-
Moment-preserving thresholding动能保持法[2]：W. Tsai, “Moment-preserving thresholding: a new approach,” Comput.Vision Graphics Image Process., vol. 29, pp. 377-393, 1985.
-
Kittler最小错误分类法[3]：Kittler, J & Illingworth, J (1986), "Minimum error thresholding", Pattern Recognition 19: 41-47
-
ISODATA(也叫做intermeans法）[4]：Ridler, TW & Calvard, S (1978), "Picture thresholding using an iterative selection method", IEEE Transactions on Systems, Man and Cybernetics 8: 630-632, http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4310039
-
Yen法[5]：1) Yen J.C., Chang F.J., and Chang S. (1995) "A New Criterion for Automatic Multilevel Thresholding" IEEE Trans. on Image Processing, 4(3): 370-378 2) Sezgin M. and Sankur B. (2004) "Survey over Image Thresholding Techniques and Quantitative Performance Evaluation" Journal of Electronic Imaging, 13(1): 146-165
-
Garbor Based[6]：Tao X, Xu D, Zhang Z T, et al. ”Weak scratch detection and defect classification methods for a large-aperture optical element”. Optics Communications, 2017, 387: 390-400.