TY - GEN
T1 - Automatic Detection Method of Target Axis and Mark Ring Based on Edge Optimization and Line Screening
AU - Yang, Xinrui
AU - Chen, Junbiao
AU - Gao, Mingwei
AU - Cao, Jing
AU - Wang, Kai
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024
Y1 - 2024
N2 - Aiming at the problem that traditional target contour and central axis detection algorithms are easily affected by factors such as non-uniform backgrounds, incomplete target imaging, and tail flame interference, which can lead to deviations in the central axis and mass center, an automatic detection method for the object central axis and mark rings based on edge optimization and line screening is proposed. Firstly, the frame difference algorithm is combined with canny algorithm to initially locate the ROI region of the moving target. Then, an improved edge detection algorithm based on the OTSU method is used to automatically extract the effective edges. Firstly, the frame difference algorithm is combined with the canny algorithm to initially locate the ROI region of the moving target. Then, an improved edge detection algorithm based on the OTSU algorithm is used to automatically extract the effective edges. After that, the target attitude interval and length threshold are calculated using EDLines algorithm, and the central axis is selected based on hough transform and line screening criteria. Finally, the mark rings are detected through RGB image segmentation and morphological operations, and their intersections with the central axis are used as the feature points. The experimental results show that the proposed method has obvious advantages over the traditional algorithms in engineering application, detection efficiency and accuracy.
AB - Aiming at the problem that traditional target contour and central axis detection algorithms are easily affected by factors such as non-uniform backgrounds, incomplete target imaging, and tail flame interference, which can lead to deviations in the central axis and mass center, an automatic detection method for the object central axis and mark rings based on edge optimization and line screening is proposed. Firstly, the frame difference algorithm is combined with canny algorithm to initially locate the ROI region of the moving target. Then, an improved edge detection algorithm based on the OTSU method is used to automatically extract the effective edges. Firstly, the frame difference algorithm is combined with the canny algorithm to initially locate the ROI region of the moving target. Then, an improved edge detection algorithm based on the OTSU algorithm is used to automatically extract the effective edges. After that, the target attitude interval and length threshold are calculated using EDLines algorithm, and the central axis is selected based on hough transform and line screening criteria. Finally, the mark rings are detected through RGB image segmentation and morphological operations, and their intersections with the central axis are used as the feature points. The experimental results show that the proposed method has obvious advantages over the traditional algorithms in engineering application, detection efficiency and accuracy.
KW - canny
KW - central axis detection
KW - hough transformation
KW - image segmentation
KW - ROI
UR - http://www.scopus.com/inward/record.url?scp=105004578378&partnerID=8YFLogxK
U2 - 10.1109/ICICML63543.2024.10958023
DO - 10.1109/ICICML63543.2024.10958023
M3 - Conference contribution
AN - SCOPUS:105004578378
T3 - 2024 International Conference on Image Processing, Computer Vision and Machine Learning, ICICML 2024
SP - 1151
EP - 1156
BT - 2024 International Conference on Image Processing, Computer Vision and Machine Learning, ICICML 2024
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 3rd International Conference on Image Processing, Computer Vision and Machine Learning, ICICML 2024
Y2 - 22 November 2024 through 24 November 2024
ER -