The Research on Fastener Image Skew Detection


Article Preview

When image digitalization device translate fastener into fastener image, the fastener image is always skewed to some extent, which will result in failure of the subsequent processing. The correction of the fastener image is the important step in its automatic recognition. In order to overcome the heavy computing burdens of hough transform, a new usage of hough transform is introduced in this paper. The algorithm works by first doing certain process on input image, the close edge of the image is gotten. Then a two-stage Hough transform algorithm is applied to the image to calculate the angle of the main edge line. This angle is thought as the declining angle of the fastener image. Lastly, the orientated image is rectified using rotation method. This algorithm is validated through experimental results.



Edited by:

Dongye Sun, Wen-Pei Sung and Ran Chen




H. R. Xu and W. H. Lu, "The Research on Fastener Image Skew Detection", Applied Mechanics and Materials, Vols. 121-126, pp. 4224-4228, 2012

Online since:

October 2011




[1] Li wen-ju, Liang de-qun: a method of quality degradation for the character segmentation of the license plate character. Computer Aided Design and Computer Graphics, Vol. 16(2004).

[2] He bin, Ma tian-yu. Wang yun-jian: Visual C + + Digital Image Processing(Posts & Telecom Press, Bei jing 2002), in Chinese.

[3] He xi-ping, Li yun-feng, Zhu qing-sheng: an efficient algorithm for automatic skew-correction of color document image. journal of image and graphics, Vol. 11(2006).

[4] Theo Gevers, Arnold W.M. Color-based Object Recognition. Pattern Recognition. (1999).

[5] Wang zhi, he sai-xian: adaptive edge detection method based on canny. journal of image and graphics, Vol. 9(2004).