A research on extraction of meter line of steel tape image based on Gabor transformation

In order to raise the robustly of the automatic verification system of steel measuring tapes, a method on extracting the meter lines on steel measuring tapes image based on Gabor transformation is proposed in this paper. This method makes full use of the frequency and direction sensitivity of Gabor filters. Through the Gabor filtering, the particular direction and frequency line structures are filtered out. Based on these line structures, positions of the meter lines in the images can be located and an ROI with the lines in it can be created automatically. Furthermore, through the RANSAC calculation to fit the double edges in the ROI, this method is able to distinguish the broken and stain lines to improve the line detect accuracy and the robustness of the image processing system. Down sampling is used to improve the efficiency of the method. An experiment has been designed to test and verify the method. By standard steel measuring tapes and dual-frequency laser interferometer, the equivalence between pixel and the physical length can be calculated. The moving distance can be calculated with the equivalence and line marker center by the method proposed in the paper. Comparing the calculated value of moving distance with reference value of moving distance measured by dual-frequency laser interferometer, the method proposed in this paper has been proof with accuracy. With the capability of solving stain, complex background, uneven illumination problems, this method can raise the degree of automation of steel measuring tapes verification system.