广东工业大学学报 ›› 2020, Vol. 37 ›› Issue (06): 63-70.doi: 10.12052/gdutxb.200006

• • 上一篇    下一篇

基于差分响应图的无监督特征点检测网络

林璟怡, 李东, 胡晓瑞   

  1. 广东工业大学 自动化学院，广东 广州 510006
  • 收稿日期:2020-01-07 出版日期:2020-11-02 发布日期:2020-11-02
  • 通信作者: 李东(1983–)，男，副教授，博士，硕士生导师，主要研究方向为模式识别、机器学习、人脸识别和机器视觉，E-mail：dong.li@gdut.edu.cn E-mail:dong.li@gdut.edu.cn
  • 作者简介:林璟怡(1994-)，男，硕士研究生，主要研究方向为图像处理、模式识别
  • 基金资助:
    国家自然科学基金资助项目（61503084）；广东省自然科学基金资助项目（2016A030310348）

An Unsupervised Feature Point Detection Network Based on Difference Response Graph

Lin Jing-yi, Li Dong, Hu Xiao-rui   

  1. School of Automation, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2020-01-07 Online:2020-11-02 Published:2020-11-02

HTML

PDF (PC)

2568

摘要： 为了突破基于人工设计的特征点检测器的性能限制，提出了一种新的数据驱动的基于差分特征响应图的无监督特征点检测网络。该网络使用不同尺度的卷积核计算差分输出，利用差分响应图的绝对值大小筛选出大量特征点，并评判这些特征点重要性程度。同时采用旋转、光照、模糊等多种图像变换训练检测器，获得相应特征不变性，使得该网络更适用于小规模数据集训练。通过在3个常用数据集上进行实验，并与现有经典算法进行定性与定量的对比分析，结果表明，基于差分响应图的无监督特征点检测网络能较好地完成特征点检测任务，所需训练时间更短、边缘定位更准确、数据集规模依赖性更低，优化特征点检测性能。

关键词: 特征点, 无监督网络, 检测

Abstract: In order to break through the performance limitations of hand-craft feature point detectors, a new data-driven unsupervised feature point detection network based on differential feature response graphs is proposed. The network calculates the differential output using convolution kernels of different scales, uses the absolute value of the differential response map to select a large number of feature points and judges the importance of these feature points. The detector is trained by using various image transformations such as rotation, illumination, and blur to obtain corresponding feature invariance, making the network more suitable for small-scale data set training. Through experiments on three commonly used data sets, and qualitative and quantitative comparison analysis with existing classic algorithms, the experimental results show that the proposed unsupervised feature point detection network based on the differential response graph can complete the feature point detection well. The network locates object edges more accurately, requires less training time, and has less dependency on dataset size. The network has optimized and improved feature point detection performance.

Key words: feature points, unsupervised network, detection

中图分类号: 

  • TP391.4
[1] SCHMID C, MOHR R, BAUCKHAGE C. Evaluation of interest point detectors [J]. International Journal of Computer Vision, 2000, 37(2): 151-172.
[2] CANNY J. A computational approach to edge detection [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6): 679-698.
[3] TRAJKOVIM, HEDLEY M. Fast corner detection [J]. Image and Vision Computing, 1998, 16(2): 75-87.
[4] MATAS J, CHUM O, URBAN M, et al. Robust wide-baseline stereo from maximally stable extremal regions [J]. Image and Vision Computing, 2004, 22(10): 761-767.
[5] HARRIS C G, STEPHENS M. A combined corner and edge detector[C]//Proceedings of the 4th Alvey Vision Conference. Manchester: Citeseer, 1988: 147-151.
[6] LOWE D G. Distinctive image features from scale-invariant keypoints [J]. International Journal of Computer Vision, 2004, 60(2): 91-110.
[7] BAY H, TUYTELAARS T, VAN GOOL L. SURF: Speeded up robust features[C]//European Conference on Computer Vision. Heidelberg: Springer, 2006: 404-417.
[8] 陈晓宁, 赵健, 杨润丰. 基于兴趣点检测的视觉关键词提取新方法研究[J]. 现代电子技术, 2017, 40(21): 70-73.
CHENG X N, ZHAO J, YANG R F. Study on visual keywords extraction new method based on points of interest detection [J]. Modern Electronics Technique, 2017, 40(21): 70-73.
[9] 李德隆, 刘伟. 基于改进的 SIFT 特征点的双目定位[J]. 广东工业大学学报, 2017, 34(1): 90-94.
LI D L, LIU W. Object location for binocular stereo vision based on improved SIFT feature [J]. Journal of Guangdong University of Technology, 2017, 34(1): 90-94.
[10] 叶志坚, 王福龙. 一种改进的 FREAK 算法的图像特征点匹配[J]. 广东工业大学学报, 2017, 34(6): 37-42.
YE Z J, WANG F L. An improved FREAK algorithm for image feature point matching [J]. Journal of Guangdong University of Technology, 2017, 34(6): 37-42.
[11] LENC K, VEDALDI A. Learning covariant feature detectors[C]//European Conference on Computer Vision. Amsterdam: Springer, 2016: 100-117.
[12] VERDIE Y, YI K, FUA P, et al. TILDE: A temporally invariant learned detector[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 5279-5288.
[13] SAVINOV N, SEKI A, LADICKY L, et al. Quad-networks: unsupervised learning to rank for interest point detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Hawaii: IEEE, 2017: 1822-1830.
[14] ONO Y, TRULLS E, FUA P, et al. LF-Net: learning local features from images[C]//Advances in Neural Information Processing Systems. San Francisco: Margan Kaufmann, 2018: 6234-6244.
[15] DETONE D, MALISIEWICZ T, RABINOVICH A. Superpoint: self-supervised interest point detection and description[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE, 2018: 224-236.
[16] MIKOLAJCZYK K, TUYTELAARS T, SCHMID C, et al. A comparison of affine region detectors [J]. International Journal of Computer Vision, 2005, 65(1-2): 43-72.
[1] 谢国波, 林立, 林志毅, 贺笛轩, 文刚. 基于YOLOv4-MP的绝缘子爆裂缺陷检测方法[J]. 广东工业大学学报, 2023, 40(02): 15-21.
[2] 陈靖宇, 吕毅. 基于脉冲神经网络的冷链制冷机结霜检测方法[J]. 广东工业大学学报, 2023, 40(01): 29-38.
[3] 李冀晖, 黄锦敬, 崔春强. 基于延迟时间模型的船舶预防性机会维修策略[J]. 广东工业大学学报, 2023, 40(01): 92-99,136.
[4] 刘信宏, 苏成悦, 陈静, 徐胜, 罗文骏, 李艺洪, 刘拔. 高分辨率桥梁裂缝图像实时检测[J]. 广东工业大学学报, 2022, 39(06): 73-79.
[5] 刘高勇, 黄靖钊, 艾丹祥. 融合立场检测和主题挖掘的突发公共事件网络舆情演化研究[J]. 广东工业大学学报, 2022, 39(03): 32-40,48.
[6] 杨俊松, 钟颖颖, 余倩. IrO2@MnO2纳米酶的制备及用于抗坏血酸的检测[J]. 广东工业大学学报, 2022, 39(02): 130-136.
[7] 杨积升, 章云, 李东. 点云目标检测残差投票网络[J]. 广东工业大学学报, 2022, 39(01): 56-62.
[8] 杨运龙, 梁路, 滕少华. 一种双路网络语义分割模型[J]. 广东工业大学学报, 2022, 39(01): 63-70.
[9] 谢胜利, 廖文建, 白玉磊, 梁勇, 董博. 相衬光学相干层析在无损检测领域的应用[J]. 广东工业大学学报, 2021, 38(06): 20-28.
[10] 梁轰, 冯丽, 徐方鑫, 李光程, 周郭许. 算力网络中一种新颖的看门狗故障检测协议[J]. 广东工业大学学报, 2021, 38(06): 35-46.
[11] 张国生, 冯广, 李东. 基于姿态表示的航空影像旋转目标检测网络[J]. 广东工业大学学报, 2021, 38(05): 40-47.
[12] 黄剑航, 王振友. 基于特征融合的深度学习目标检测算法研究[J]. 广东工业大学学报, 2021, 38(04): 52-58.
[13] 汝少楠, 何元烈, 叶星余. 基于稀疏直接法闭环检测定位的视觉里程计[J]. 广东工业大学学报, 2021, 38(03): 48-54.
[14] 滕少华, 陈成, 霍颖翔. 小样本纠错的多层入侵检测分类研究[J]. 广东工业大学学报, 2020, 37(03): 9-16.
[15] 钟映春, 孙思语, 吕帅, 罗志勇, 熊勇良, 何惠清. 铁塔航拍图像中鸟巢的YOLOv3识别研究[J]. 广东工业大学学报, 2020, 37(03): 42-48.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发