Research of Adaptive Extracting Algorithm Facing Blade Crack Image

Abstract:

RT（X-Radiography Testing） is a very import inspection method for detecting engine blade crack, and the defect is identified by person distinguishing the film for traditional RT. The traditional method is involved in high mistake and missed examining rate and low efficiency. It is great significance to blade integrity and flight safety that the crack can be quickly distinguished. In the paper, adopting a adaptive segmenting algorithm based on OTSU to quickly distinguish the crack on the blade image, and it can definite the threshold facing different gray image, and can reliably inspect the cracks which detection rate was 98%. The recognition accuracy for crack can reach 0.16mm Simulation result verify that the method is very effective on adaptively segmenting the blade image according to the threshold for crack inspection

• Periodical: Applied Mechanics and Materials (Volumes 135-136)
• Pages: 732-738
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.135-136.732
• Online since: October 2011
• Authors: Li Min Wen, Chang Bao Wen, Yan Ming Li
• Keywords: Adaptive Segmentation, Blade, Crack, Otsu, RT

References

[1] DING Peng, LI Chang you, MA Qi shuang, Wavelet based fault detection of aeronautic engine vanes by borescope, Transactions of Beijing University of Aeronautics and Astronautics, Vol 32, Dec 2006，pp.435-1438.

[2] ZHOU Zheng gan, TENG Sheng hua, JIANG Wei. Evaluation on Method of X-ray Inspection for Welding, Transactions of the China Welding Institution, Vol 23, Feb, 2002, pp.85-88.

[3] LI Liao liao, DENG Shan xi, DING Xing hao. Binarization Algorithm Based on Image Partition derived from OTSU, CONTROL & AUTOMATION, Vol 21, Aug, 2005, pp.76-78.

[4] ZHOU Zheng gan, DU Yuan yuan, Techniques of Automatic Inspection for Defects Based on multi X-ray images, CHINESE JOURNAL OF MECHANICAL ENGINEERING, Vol 42 , Mar, 2006, pp.73-76.

[5] LUO Xi ping, TIAN Jie. A Survey on Image Segmentation Methods, Pattern Recognition and Artificial Intelligence, Vol 12, 1998, pp.300-309.

[6] Otsu .N.A. Threshold Selection Method from Gray Level Histogram[J]. IEEETrans, 1979, SMC9(1): 62-66.

[7] Zhao-chang DAI Yao-dong ZHOU Zheng-dong. Mamm- graphic mass segmentation based on Otsu multi-threshold and watershed algorithm[J]. Computer Applications. 2008, 12(28): 198-200.

[8] WU Cheng mao, TIAN Xiao ping, TAN Tie niu. Fast Iterative Algorithm for Two-Dimensional Otsu Thresholding Method, Pattern Recognition and Artificial Intelligence, Vol 21, Jun, 2008, pp.746-757.