《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (1): 167-174.DOI: 10.11772/j.issn.1001-9081.2023010037
所属专题: 人工智能
收稿日期:
2023-01-15
修回日期:
2023-04-13
接受日期:
2023-04-14
发布日期:
2023-06-06
出版日期:
2024-01-10
通讯作者:
邢艳
作者简介:
黄睿(1987—),男,宁夏中卫人,讲师,博士,CCF会员,主要研究方向:计算机视觉、图像处理、机器学习;基金资助:
Rui HUANG1, Chaoqun ZHANG1, Xuyi CHENG1, Yan XING1(), Bao ZHANG2
Received:
2023-01-15
Revised:
2023-04-13
Accepted:
2023-04-14
Online:
2023-06-06
Published:
2024-01-10
Contact:
Yan XING
About author:
HUANG Rui, born in 1987, Ph. D., lecturer. His research interests include computer vision, image processing, machine learning.Supported by:
摘要:
当前基于深度学习的实例检测方法在进行发动机叶片分割时,由于缺少带标注的发动机叶片数据,导致无法充分训练网络模型,仅得到次优的分割结果。为了提升航空发动机叶片实例分割精度,提出了不完整实例引导的航空发动机叶片实例分割方法,通过结合已有的实例分割方法和交互式分割方法,可得到较好的发动机叶片分割结果。首先,使用少量标注数据训练实例分割网络,得到发动机叶片的初步分割结果;其次,将检测到的单个叶片分为前景和背景两部分,通过选择前景种子点和背景种子点,利用交互式分割方法的思想,产生完整的单个叶片的分割结果;依次处理完所有的叶片后,将结果合并得到最终的发动机叶片实例分割结果。使用72张图像训练基于稀疏实例激活图的实时实例分割方法(SparseInst)产生初始的实例分割结果,在56张图像上进行测试。所提方法的全类平均准确率(mAP)比SparseInst的全类平均准确率高5.1个百分点;且它的mAP结果均优于当前流行的实例分割方法MASK R-CNN(MASK Region based Convolutional Neural Network)、YOLACT (You Only Look At CoefficienTs)、BMASK-RCNN (Boundary-preserving MASK R-CNN)。
中图分类号:
黄睿, 张超群, 成旭毅, 邢艳, 张宝. 不完整实例引导的航空发动机叶片实例分割[J]. 计算机应用, 2024, 44(1): 167-174.
Rui HUANG, Chaoqun ZHANG, Xuyi CHENG, Yan XING, Bao ZHANG. Incomplete instance guided aeroengine blade instance segmentation[J]. Journal of Computer Applications, 2024, 44(1): 167-174.
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
BMASK R-CNN[ | 0.393 | 0.695 | 0.381 |
YOLACT[ | 0.325 | 0.571 | 0.318 |
MASK R-CNN[ | 0.475 | 0.736 | 0.501 |
SparseInst[ | 0.579 | 0.809 | 0.618 |
本文方法 | 0.630 | 0.855 | 0.725 |
表1 不同实例分割方法的航空发动机叶片分割结果量化比较
Tab. 1 Quantitative comparison of segmentation results of aeroengine blade among different instance segmentation methods
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
BMASK R-CNN[ | 0.393 | 0.695 | 0.381 |
YOLACT[ | 0.325 | 0.571 | 0.318 |
MASK R-CNN[ | 0.475 | 0.736 | 0.501 |
SparseInst[ | 0.579 | 0.809 | 0.618 |
本文方法 | 0.630 | 0.855 | 0.725 |
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
只使用背景点 | 0.098 | 0.112 | 0.100 |
只使用前景点 | 0.509 | 0.710 | 0.572 |
本文方法 | 0.630 | 0.855 | 0.725 |
表2 前景点与背景点对实例分割结果的影响
Tab. 2 Effect of foreground and background points on instance segmentation result
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
只使用背景点 | 0.098 | 0.112 | 0.100 |
只使用前景点 | 0.509 | 0.710 | 0.572 |
本文方法 | 0.630 | 0.855 | 0.725 |
K | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
3 | 0.518 | 0.706 | 0.589 |
4 | 0.630 | 0.855 | 0.725 |
5 | 0.629 | 0.853 | 0.725 |
表3 前景点的数量对实例分割结果的影响
Tab. 3 Effect of number of foreground points on instance segmentation result
K | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
3 | 0.518 | 0.706 | 0.589 |
4 | 0.630 | 0.855 | 0.725 |
5 | 0.629 | 0.853 | 0.725 |
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
MASK R-CNN[ | 0.475 | 0.736 | 0.501 |
MASK R-CNN[ | 0.546 | 0.761 | 0.670 |
表4 MASK R-CNN分割结果与再使用本文方法得到的分割结果的指标对比
Tab. 4 Comparison of segmentation metrics for results of MASK R-CNN and MASK R-CNN +proposed method
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
MASK R-CNN[ | 0.475 | 0.736 | 0.501 |
MASK R-CNN[ | 0.546 | 0.761 | 0.670 |
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
BMASK R-CNN[ | 0.393 | 0.695 | 0.381 |
BMASK R-CNN[ | 0.424 | 0.732 | 0.415 |
表5 BMASK R-CNN分割结果与再使用本文方法得到的分割结果的指标对比
Tab. 5 Comparison of segmentation metrics for results of BMASK R-CNN and BMASK R-CNN +proposed method
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
BMASK R-CNN[ | 0.393 | 0.695 | 0.381 |
BMASK R-CNN[ | 0.424 | 0.732 | 0.415 |
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
YOLACT[ | 0.325 | 0.571 | 0.318 |
YOLACT[ | 0.390 | 0.605 | 0.423 |
表6 YOLACT分割结果与再使用本文方法得到的分割结果的指标对比
Tab. 6 Comparison of segmentation metrics for results of YOLACT and YOLACT +proposed method
方法 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|
YOLACT[ | 0.325 | 0.571 | 0.318 |
YOLACT[ | 0.390 | 0.605 | 0.423 |
T1 | T2 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|---|
0.6 | 0.7 | 0.630 | 0.855 | 0.725 |
0.7 | 0.8 | 0.618 | 0.858 | 0.693 |
0.8 | 0.6 | 0.624 | 0.855 | 0.705 |
0.9 | 0.6 | 0.624 | 0.855 | 0.705 |
0.6 | 0.9 | 0.573 | 0.856 | 0.598 |
表7 阈值T1、T2对实例分割结果的影响
Tab. 7 Influence of threshold T1 and T2 on instance segmentation results
T1 | T2 | mAP | AP(0.5) | AP(0.75) |
---|---|---|---|---|
0.6 | 0.7 | 0.630 | 0.855 | 0.725 |
0.7 | 0.8 | 0.618 | 0.858 | 0.693 |
0.8 | 0.6 | 0.624 | 0.855 | 0.705 |
0.9 | 0.6 | 0.624 | 0.855 | 0.705 |
0.6 | 0.9 | 0.573 | 0.856 | 0.598 |
1 | 张勇.孔探工作在发动机维护中的重要性[J].航空维修与工程, 2004(1): 24-25. 10.3969/j.issn.1672-0989.2004.01.010 |
ZHANG Y. The importance of borescope inspection in aircraft engine maintenance [J]. Aviation Maintenance & Engineering, 2004(1): 24-25. 10.3969/j.issn.1672-0989.2004.01.010 | |
2 | HE K, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN [C]// Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2980-2988. 10.1109/iccv.2017.322 |
3 | LIN T-Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: Common objects in context [C]// Proceedings of the 2014 European Conference on Computer Vision. Cham: Springer, 2014: 740-755. 10.1007/978-3-319-10602-1_48 |
4 | CHENG T, WANG X, HUANG L, et al. Boundary-preserving Mask R-CNN [C]// Proceedings of the 2020 European Conference on Computer Vision. Cham: Springer, 2020: 660-676. 10.1007/978-3-030-58568-6_39 |
5 | CHEN K, PANG J, WANG J, et al. Hybrid task cascade for instance segmentation [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 4969-4978. 10.1109/cvpr.2019.00511 |
6 | SOFIIUK K, PETROV I, BARINOVA O, et al. f-BRS: Rethinking backpropagating refinement for interactive segmentation [C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2020: 8620-8629. 10.1109/cvpr42600.2020.00865 |
7 | CHENG T, WANG X, CHEN S, et al. Sparse instance activation for real-time instance segmentation [C]// Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 4423-4432. 10.1109/cvpr52688.2022.00439 |
8 | 旷可嘉.深度学习及其在航空发动机缺陷检测中的应用研究[D].广州:华南理工大学, 2017. |
KUANG K J. Research on deep learning and its application on the defects detection for aero engine [D]. Guangzhou: South China University of Technology, 2017. | |
9 | KIM Y-H, LEE J-R. Videoscope-based inspection of turbofan engine blades using convolutional neural networks and image processing [J]. Structural Health Monitoring, 2019, 18(5/6): 2020-2039. 10.1177/1475921719830328 |
10 | 黄睿,段博坤,陈望,等.检测器和分类器级联的飞机发动机损伤检测[J].中国图象图形学报, 2022, 27(11): 3232-3242. 10.11834/jig.210734 |
HUANG R, DUAN B K, CHEN W, et al. Detector and classifier cascaded aero-engine detection method [J]. Journal of Image and Graphics, 2022, 27(11): 3232-3242. 10.11834/jig.210734 | |
11 | LI D, LI Y, XIE Q, et al. Tiny defect detection in high-resolution aero-engine blade images via a coarse-to-fine framework [J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-12. 10.1109/tim.2021.3062175 |
12 | ZHANG D, ZENG N, LIN L. Detection of blades damages in aero engine [C]// Proceedings of the 2020 Chinese Automation Congress. Piscataway: IEEE, 2020: 6129-6134. 10.1109/cac51589.2020.9327202 |
13 | BOLYA D, ZHOU C, XIAO F, et al. YOLACT: Real-time instance segmentation [C]// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2019: 9156-9165. 10.1109/iccv.2019.00925 |
14 | TIAN Z, SHEN C, CHEN H. Conditional convolutions for instance segmentation [C]// Proceedings of the 2020 European Conference on Computer Vision. Cham: Springer, 2020: 282-298. 10.1007/978-3-030-58452-8_17 |
15 | LIU S, QI L, QIN H, et al. Path aggregation network for instance segmentation [C]// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 8759-8768. 10.1109/cvpr.2018.00913 |
16 | LIN G, MILAN A, SHEN C, et al. RefineNet: Multi-path refinement networks for high-resolution semantic segmentation [C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, 2017: 5168-5177. 10.1109/cvpr.2017.549 |
17 | WANG X, KONG T, SHEN C, et al. SOLO: Segmenting objects by locations [C]// Proceedings of the 2020 European Conference on Computer Vision. Cham: Springer, 2020: 649-665. 10.1007/978-3-030-58523-5_38 |
18 | 樊玮,李晨炫,邢艳,等.航空发动机损伤图像的二分类到多分类递进式检测网络[J].计算机应用, 2021, 41(8): 2352-2357. 10.11772/j.issn.1001-9081.2020101575 |
FAN W, LI C X, XING Y, et al. Binary classification to multiple classification progressive detection network for aero-engine damage images [J]. Journal of Computer Applications, 2021, 41(8): 2352-2357. 10.11772/j.issn.1001-9081.2020101575 | |
19 | CHEN L-C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 40(4): 834-848. 10.1109/tpami.2017.2699184 |
20 | 马佳良,陈斌,孙晓飞.基于改进的Faster R-CNN的通用目标检测框架[J].计算机应用, 2021, 41(9): 2712-2719. 10.11772/j.issn.1001-9081.2020111852 |
MA J L, CHEN B, SUN X F. General object detection framework based on improved Faster R-CNN [J]. Journal of Computer Applications, 2021, 41(9): 2712-2719. 10.11772/j.issn.1001-9081.2020111852 | |
21 | ROTHER C, KOLMOGOROV V, BLAKE A. “GrabCut” interactive foreground extraction using iterated graph cuts [J]. ACM Transactions on Graphics, 2004, 23(3): 309-314. 10.1145/1015706.1015720 |
22 | CHEN X, ZHAO Z, ZHANG Y, et al. FocalClick: Towards practical interactive image segmentation [C]// Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 1290-1299. 10.1109/cvpr52688.2022.00136 |
23 | SOFIIUK K, PETROV I A, KONUSHIN A. Reviving iterative training with mask guidance for interactive segmentation [C]// Proceedings of the 2022 IEEE International Conference on Image Processing. Piscataway: IEEE, 2022: 3141-3145. 10.1109/icip46576.2022.9897365 |
24 | XU N, PRICE B, COHEN S, et al. Deep interactive object selection [C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 373-381. 10.1109/cvpr.2016.47 |
25 | 兰红,闵乐泉.多阈值优化交互式分割算法及其在医学图像中的应用[J].计算机应用, 2013, 33(5): 1435-1438, 1475. 10.3724/sp.j.1087.2013.01435 |
LAN H, MIN L Q. Interactive segmentation algorithm optimized by multi-threshold with application in medical images [J]. Journal of Computer Applications, 2013, 33(5): 1435-1438, 1475. 10.3724/sp.j.1087.2013.01435 | |
26 | CHEN X, ZHAO Z, YU F, et al. Conditional diffusion for interactive segmentation [C]// Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 7325-7334. 10.1109/iccv48922.2021.00725 |
27 | LIN Z, ZHANG Z, CHEN L-Z, et al. Interactive image segmentation with first click attention [C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2020: 13336-13345. 10.1109/cvpr42600.2020.01335 |
28 | ACUNA D, LING H, KAR A, et al. Efficient interactive annotation of segmentation datasets with polygon-RNN++ [C]// Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 859-868. 10.1109/cvpr.2018.00096 |
29 | SHAMOS M I. Computational geometry [D]. New Haven: Yale University, 1978. |
30 | SUN K, XIAO B, LIU D, et al. Deep high-resolution representation learning for human pose estimation [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 5686-5696. 10.1109/cvpr.2019.00584 |
31 | GUPTA A, DOLLÁR P, GIRSHICK R. LVIS: A dataset for large vocabulary instance segmentation [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 5351-5359. 10.1109/cvpr.2019.00550 |
[1] | 孙逊, 冯睿锋, 陈彦如. 基于深度与实例分割融合的单目3D目标检测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2208-2215. |
[2] | 廖存燚, 郑毅, 刘玮瑾, 于欢, 刘守印. 自动驾驶环境感知多任务去耦-融合算法[J]. 《计算机应用》唯一官方网站, 2024, 44(2): 424-431. |
[3] | 赵元龙, 单玉刚, 袁杰, 赵康迪. 基于实例分割与毕达哥拉斯模糊决策的目标跟踪[J]. 《计算机应用》唯一官方网站, 2023, 43(6): 1930-1937. |
[4] | 高榕, 沈加伟, 邵雄凯, 吴歆韵. 基于Fastformer和自监督对比学习的实例分割算法[J]. 《计算机应用》唯一官方网站, 2023, 43(4): 1062-1070. |
[5] | 肖田邹子, 周小博, 罗欣, 唐其鹏. 动态环境下结合实例分割与聚类的鲁棒RGB-D SLAM系统[J]. 《计算机应用》唯一官方网站, 2023, 43(4): 1220-1225. |
[6] | 刘月峰, 张小燕, 郭威, 边浩东, 何滢婕. 基于优化混合模型的航空发动机剩余寿命预测方法[J]. 《计算机应用》唯一官方网站, 2022, 42(9): 2960-2968. |
[7] | 马佳良, 陈斌, 孙晓飞. 基于改进的Faster R-CNN的通用目标检测框架[J]. 计算机应用, 2021, 41(9): 2712-2719. |
[8] | 樊玮, 李晨炫, 邢艳, 黄睿, 彭洪健. 航空发动机损伤图像的二分类到多分类递进式检测网络[J]. 计算机应用, 2021, 41(8): 2352-2357. |
[9] | 冯兴杰, 张天泽. 基于分组卷积进行特征融合的全景分割算法[J]. 计算机应用, 2021, 41(7): 2054-2061. |
[10] | 章悦, 张亮, 谢非, 杨嘉乐, 张瑞, 刘益剑. 基于实例分割模型优化的道路抛洒物检测算法[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3228-3233. |
[11] | 田锦, 袁家政, 刘宏哲. 基于实例分割的车道线检测及自适应拟合算法[J]. 计算机应用, 2020, 40(7): 1932-1937. |
[12] | 冯涛, 陈斌, 张跃飞. 基于改进的Mask R-CNN的染色体图像分割框架[J]. 计算机应用, 2020, 40(11): 3332-3339. |
[13] | 黄辉先 任科明 李燕 庄选. 基于渐消卡尔曼滤波算法的航空发动机参数估计方法[J]. 计算机应用, 2013, 33(10): 2993-2995. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||