Optics and Precision Engineering, Volume. 32, Issue 1, 125(2024)
Hole feature detection for aircraft parts by integrating visual saliency and group decision making
Figures & Tables(16)
Fig. 1. Original hole images and detection result of classical FT method
Fig. 2. Saliency detection results based on preliminary improved FT method
Fig. 3. Saliency detection results based on final improved FT method
Fig. 5. Detection results of novel mathematical morphological contour extraction method
Fig. 9. Hole feature detection results under uneven lighting and hole defects
Fig. 11. Impact of different salt-and-pepper noise densities on detection results of hole parameters
Table 1. Pseudocode of algorithm
View tableView in ArticleTable 1. Pseudocode of algorithm
Algorithm 1 Contour_FittingInput： Cen： Center obtained by segmented fitting
Rad： Radius obtained by segmented fitting
Meanshift： Center localization algorithm
Groupdecision： Radius calculation algorithm
Output： Circle： Detected circle
1：
PreCen = PreProcess（Cen）
2：
while Te > Tsdo 3：
Point = Meanshift（PreCen）
4：
end 5：
R = FindRadius（Point， Rad）
6：
Radius = Groupdecision （R）
7：
Circle = DrawCircle（Point， Radius）
8：
return Circle
Table 2. Comparison of hole radius detection errors of various algorithms
View tableView in ArticleTable 2. Comparison of hole radius detection errors of various algorithms
组别 本文方法 H方法 D方法 1 0.007 0.015 0.029 2 0.015 0.052 0.038 3 0.011 1.001 0.976 4 0.012 0.016 2.897 平均误差 0.011 0.271 0.985
Table 3. Comparison of detection time of various algorithms
View tableView in ArticleTable 3. Comparison of detection time of various algorithms
组别 本文方法 H方法 D方法 1 0.218 0.159 0.161 2 0.238 0.204 0.234 3 0.185 0.179 0.169 4 0.302 0.228 3.673 平均时间 0.236 0.193 1.059
Tools
Get Citation
Copy Citation Text
Jiachun TIAN, Liang WANG, Biao MEI, Weidong ZHU. Hole feature detection for aircraft parts by integrating visual saliency and group decision making[J]. Optics and Precision Engineering, 2024, 32(1): 125
Paper Information
Category:
Received: Jul. 17, 2023
Accepted: --
Published Online: Jan. 23, 2024
The Author Email: Biao MEI (mechme@126.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20