Image processing and analysis algorithms for measuring log volume
Long Defan1, Fan Shangchun1, Pang Hongbing2*
1. School of Automation Science and Electrical Engineering, Beijing University of Aeronautics and Astronautics, Beijing 100083, China;
2. School of Computer Science and Technology, Beijing University of Aeronautics and Astronautics, Beijing 100083, China
The key image processing and analysis algorithms were investigated and designed based on operating principle of the automatic system for log volume measurement. Hough transform (HT) was applied to the image since log ends are quasi-circular. Peaks of the HT were detected in maximal balls corresponding toeach maximum skeleton point. Consequently, the image was coarsely segmented by each HT peak corresponding to one log end. The active contour model was transformed into the polar coordinates. And the maxima of image gradient were selected as candidate points of log end contour. Global minimum of the energy functional can be found using dynamic programming. With this approach, the precise contours of the log ends can be extracted. Experiment shows that the proposed algorithms are of significance.
�� ��,����һ.��������Բ��Ŀ��ͼ��ʶ���²���[J].�й�ͼ��ͼ��ѧ��,1999,4(3):202~206 Luan Xin, Zhu Tieyi. The new recognition algorithm for irregular quasi-circular object[J]. Journal of Image and Graphics,1999,4(3):202~206(in Chinese)
Wang Yaming, Huang Wenqing, Zhao Yun. Computer vision technology in log 13 inspection[J].Journal of Forestry Research.2002,13(1):67-70