A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Defect Detection of Steel Surfaces with Global Adaptive Percentile Thresholding of Gradient Image
Abstract Steel strips are used extensively for white goods, auto bodies and other purposes where surface defects are not acceptable. On-line surface inspection systems can effectively detect and classify defects and help in taking corrective actions. For detection of defects use of gradients is very popular in highlighting and subsequently segmenting areas of interest in a surface inspection system. Most of the time, segmentation by a fixed value threshold leads to unsatisfactory results. As defects can be both very small and large in size, segmentation of a gradient image based on percentile thresholding can lead to inadequate or excessive segmentation of defective regions. A global adaptive percentile thresholding of gradient image has been formulated for blister defect and water-deposit (a pseudo defect) in steel strips. The developed method adaptively changes the percentile value used for thresholding depending on the number of pixels above some specific values of gray level of the gradient image. The method is able to segment defective regions selectively preserving the characteristics of defects irrespective of the size of the defects. The developed method performs better than Otsu method of thresholding and an adaptive thresholding method based on local properties.
Defect Detection of Steel Surfaces with Global Adaptive Percentile Thresholding of Gradient Image
Abstract Steel strips are used extensively for white goods, auto bodies and other purposes where surface defects are not acceptable. On-line surface inspection systems can effectively detect and classify defects and help in taking corrective actions. For detection of defects use of gradients is very popular in highlighting and subsequently segmenting areas of interest in a surface inspection system. Most of the time, segmentation by a fixed value threshold leads to unsatisfactory results. As defects can be both very small and large in size, segmentation of a gradient image based on percentile thresholding can lead to inadequate or excessive segmentation of defective regions. A global adaptive percentile thresholding of gradient image has been formulated for blister defect and water-deposit (a pseudo defect) in steel strips. The developed method adaptively changes the percentile value used for thresholding depending on the number of pixels above some specific values of gray level of the gradient image. The method is able to segment defective regions selectively preserving the characteristics of defects irrespective of the size of the defects. The developed method performs better than Otsu method of thresholding and an adaptive thresholding method based on local properties.
Defect Detection of Steel Surfaces with Global Adaptive Percentile Thresholding of Gradient Image
Neogi, Nirbhar (author) / Mohanta, Dusmanta K. (author) / Dutta, Pranab K. (author)
Journal of The Institution of Engineers (India): Series B ; 98 ; 557-565
2017-11-22
9 pages
Article (Journal)
Electronic Resource
English
Defect Detection of Steel Surfaces with Global Adaptive Percentile Thresholding of Gradient Image
| Springer Verlag | 2017
Line Defect Detection in TFT-LCD Using Directional Filter Bank and Adaptive Multilevel Thresholding
| British Library Online Contents | 2004
An Adaptive Thresholding Multiple Classifiers System for Remote Sensing Image Classification
| Online Contents | 2009
Estimating Percentile Speeds from Maximum Operating Speed Frontier
| British Library Online Contents | 2014