The Study of Counting the Strapping Steel Pipes Based on Image Recognition

Abstract:

The accurate count has been realized and the expected results have been acquired by the image of the end of steel pipes recognized, which has a good market prospect and economic profit. The mathematical morphology was used to preprocess image and inspect the edge of the target. An improved Hough Transformation was applied to recognize inner diameter of steel pipes, which improved the detection rate. The Hu invariant moments described in the area was used to recognize inner diameter of steel pipes, which realized the pipe sum accurately. The method to count points was used to realize the sum of pipes according to the condition that the inner diameter circus of binary pipe image was corroded more times came to change into small area.

• Periodical: Advanced Materials Research (Volumes 490-495)
• Pages: 3831-3835
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.490-495.3831
• Online since: March 2012
• Authors: Xue Hua Li, Hui Min Liang, Zhi Kai Cao, Zhong Yun Jiang, Long Chen
• Keywords: Circular Steel Pipe, Count, Hough Transformation, Hu Invariant Moments, Image Recognition

References

[1] Qiuqi Ruan: Digital Image Processing, Beijing: Electronics Industry Press, 2001. In Chinese.

[2] Yujin Zhang: Image Processing and Analysis, Beijing: Tsinghua University Press, 2003. In Chinese.

[3] Yinan Sun, Weijun Liu, Yuechao Wang, Yipeng Sun: A Method for Circle Detection Using Modified Hough Transform. Computer Engineering and Applications, 2003(20), pp.35-37. In Chinese.

[4] Linxia Liu, Jie Chen, Lihua Dou: Invariant Moments and Their Application to Target Recognition. Fire Control & Command, 2003(2). pp.13-15. In Chinese.