Pattern recognition based on the fusion technique for counting of bundled bar steel

The counting of the bundled bar steel is an unsolved problem. To solve this problem,a pattern recognition approach which is based on the data fusion technique is presented in this paper. At first, the end image of the bundled bar steel is acquired with CCD and segmented with the threshold and edgebased methods. Then the segmented images are mixed, the adhesions are gotten rid of and the number of bar steel is counted. Experimental result shows that this technique is effective in improving the accuracy of counting.