Improved Infrared Image Edge Detection Algorithm Based on DexiNed
-
摘要: 相对于可见光图像边缘检测，目前针对红外图像边缘检测的研究较少，且大多基于传统方法，如边缘检测算子、数学形态学等，其本质上都是只考虑红外图像局部的急剧变化来检测边缘，因而始终受限于低层次特征。本文提出了一种基于深度学习的红外图像边缘检测算法，在DexiNed（Dense Extreme Inception Network for Edge Detection）的基础上，缩减了网络规模，并在损失函数中引入了图像级的差异，精心设置了损失函数的参数，进而优化了网络性能。此外，还通过调整自然图像边缘检测数据集来近似模拟红外图像边缘检测数据集，对改进后的模型进行训练，进一步提高了网络对红外图像中边缘信息的提取能力。定性评价结果表明，本文方法提取的红外图像边缘定位准确、层次清晰、细节丰富、贴合人眼视觉，使用了SSIM（Structural Similarity Index Measure）和FSIM（Feature Similarity Index Measure）指标的定量评价结果进一步体现了本文方法相比于其他方法的优越性。Abstract: Compared with optical image edge detection, there are fewer studies on infrared image edge detection, and most of them are based on traditional methods, such as edge detection operators and mathematical morphology. In essence, they only consider the sharp local changes of infrared images to detect edges, so they are always limited by low-level features. In this paper, an infrared image edge detection algorithm based on deep learning is proposed. Based on the dense extreme inception network for edge detection (DexiNed), the network capacity is reduced by removing the last main block, the image level difference is introduced into the loss function, and the parameters of the loss function are carefully set to optimize the network performance. In addition, by adjusting the natural image edge detection dataset to approximate the infrared image edge detection dataset, the improved model was trained to enhance the edge detection ability. The qualitative evaluation results show that the edge of the infrared image extracted by our method is accurate, precise, rich in detail, and fits human vision. A quantitative evaluation using the structural similarity indexmatrix (SSIM) and feature similarity indexmatrix (FSIM) indexes further shows the advantages of our method compared with other existing methods.
-
Keywords:
- infrared image /
- edge detection /
- deep learning
-
-
![]()
图 1 几种图像边缘检测方法提取红外图像边缘的效果对比：(a)红外图像（来源于FLIR红外数据集）；(b)Canny算子的边缘检测结果；(c)BDCN[12]的边缘检测结果；(d)本文方法的边缘检测结果
Figure 1. Comparison of several image edge detection methods to extract infrared image edge: (a) is an example infrared image from FLIR Thermal Dataset (www.flir.com); (b) is the result of the Canny edge detector; (c) is the result of BDCN[12]; (d) is the result of our method
![]()
图 2 DexiNed网络结构和精简后的网络（位于虚线框中）
Figure 2. Network architecture of DexiNed[14] and simplified one(in dotted box)
![]()
图 3 可见光图像转变为模拟红外图像的过程
Figure 3. The process of transforming an optical image into a simulated infrared image
![]()
图 4 测试结果对比
Figure 4. Comparison of test results
Infrared image Result 1 Result 2 Result 3 ![]()
图 5 在IR-BIPED数据集上训练前后的网络测试结果对比
Figure 5. Network test before and after training on IR-BIPED dataset
![]()
图 6 不同边缘检测方法的结果对比
Figure 6. Results of different methods
Image-1 Image-2 Image-3 Image-4 Infrared image Enhanced Canny[3] Improved ant colony [5] RCF[11] BDCN[12] Our method ![]()
图 7 不同方法在图像1上测试后的定量结果对比
Figure 7. Comparison of quantitative results of different methods on image1
![]()
图 8 不同方法在图像2上测试后的定量结果对比
Figure 8. Comparison of quantitative results of different methods on image2
![]()
图 9 不同方法在图像3上测试后的定量结果对比
Figure 9. Comparison of quantitative results of different methods on image3
![]()
图 10 不同方法在图像4上测试后的定量结果对比
Figure 10. Comparison of quantitative results of different methods on image4
表 1 网络准确度对比1
Table 1 Comparison of network accuracy-1
Baseline Simplified epoch 1 0.8991782 0.8991683 epoch 2 0.8993612 0.8995107 epoch 3 0.8993013 0.8994964 
下载: 导出CSV
表 2 网络准确度对比2
Table 2 Comparison of network accuracy-2
Baseline Simplified+designed loss epoch 1 0.8991875 0.9012373 epoch 2 0.8991702 0.9105164 epoch 3 0.8992519 0.9097796
下载: 导出CSV
-
[1] 夏清, 胡振琪, 位蓓蕾, 等. 一种新的红外热像仪图像边缘检测方法[J]. 红外与激光工程, 2014, 43(1): 318-322. DOI: 10.3969/j.issn.1007-2276.2014.01.056 XIA Qing, HU Zhenqi, WEI Beilei, et al. New edge detection method for images of infrared thermal imager[J]. Infrared and Laser Engineering, 2014, 43(1): 318-322. DOI: 10.3969/j.issn.1007-2276.2014.01.056
[2] 唐庆菊, 刘俊岩, 王扬, 等. 基于模糊C均值聚类和Canny算子的红外图像边缘识别与缺陷定量检测[J]. 红外与激光工程, 2016, 45(9): 281-285. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201609042.htm TANG Qingju, LIU Junyan, WANG Yang, et al. Infrared image edge recognition and defect quantitative determination based on the algorithm of fuzzy C-means clustering and Canny operator[J]. Infrared and Laser Engineering, 2016, 45(9): 281-285. https://www.cnki.com.cn/Article/CJFDTOTAL-HWYJ201609042.htm
[3] Minh V H, Aleef T A, Pervaiz U, et al. Smoothness-based Edge Detection using Low-SNR Camera for Robot Naviga-tion[J/OL]. arXiv preprint arXiv: 1710.01416, 2017.
[4] LIU Lisang, LIANG Fenqiang, ZHENG Jishi, et al. Ship infrared image edge detection based on an improved adaptive Canny algorithm[J]. International Journal of Distributed Sensor Networks, 2018, 14(3): 51-57. http://www.onacademic.com/detail/journal_1000040876503810_4fec.html
[5] QIAO B M, LI S, XIAO Y. Infrared image edge detection applied research based on improved ant colony algorithm[C]//Proceedings of the 2015 International conference on Applied Science and Engineering Innovation, 2015: 31.
[6] 王凤. 结合蚁群搜索与边缘检测的红外轮廓提取算法[J]. 激光与红外, 2015, 45(6): 722-727. DOI: 10.3969/j.issn.1001-5078.2015.06.025 WANG Feng. Infrared contour extraction algorithm based on ants search and edge detection[J]. Laser & Infrared, 2015, 45(6): 722-727. DOI: 10.3969/j.issn.1001-5078.2015.06.025
[7] ZHANG X, HAN R. The Application of Mathematical Morphology and Sobel Operator in Infrared Image Edge Detection[C]//International Industrial Informatics & Computer Engineering Conference, 2015: 949-954.
[8] YIN J, LU Y, GONG Z, et al. Edge Detection of High-Voltage Porcelain Insulators in Infrared Image Using Dual Parity Morphological Gradients[J]. IEEE Access, 2019, 26(7): 32728-32734. http://www.onacademic.com/detail/journal_1000042299647699_6769.html
[9] Bertasius G, SHI J, Torresani. Deep Edge: A Multi-Scale Bifurcated Deep Network for Top-Down Contour Detection[C]//Computer Vision & Pattern Recognition. IEEE, 2015: 4380-4389.
[10] XIE S, TU Z. Holistically-nested edge detection[C]//Proceedings of the IEEE international conference on computer vision, 2015: 1395-1403.
[11] LIU Y, CHENG M M, HU X, et al. Richer convolutional features for edge detection[C]//Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference, 2017: 5872-5881.
[12] HE J, ZHANG S, YANG M, et al. Bidirectional cascade network for perceptual edge detection[J/OL]. arXiv preprint arXiv: 1902.10903, 2019.
[13] DENG R, SHEN C, LIU S, et al. Learning to predictcrisp boundaries [C]//European Conference on Computer Vision, 2018: 1-17.
[14] Xavier Soria, Edgar Riba, Angel D. Sappa. Dense Extreme Inception Network: Towards a Robust CNN Model for Edge Detection[J/OL]. arXiv preprint arXiv: 1909.01955v2, 2020.
[15] 焦安波, 何淼, 罗海波. 一种改进的HED网络及其在边缘检测中的应用[J]. 红外技术, 2019, 41(1): 72-77. http://hwjs.nvir.cn/article/id/hwjs201901011 JIAO Anbo, HE Miao, LUO Haibo. Research on Significant Edge Detection of Infrared Image Based on Deep Learning[J]. Infrared Technology, 2019, 41(1): 72-77. http://hwjs.nvir.cn/article/id/hwjs201901011
[16] Mely D A, Kim J, McGill M, et al. A systematic comparison between visual cues for boundary detection[J]. Vision Research, 2016, 120: 93-107. DOI: 10.1016/j.visres.2015.11.007
[17] Grigorescu C, Petkov N, Westenberg M A. Contour detection based on nonclassical receptive field inhibition[J]. IEEE Transactions on image processing, 2003, 12(7): 729-739. DOI: 10.1109/TIP.2003.814250
[18] Martin D, Fowlkes C, Tal D, et al. A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics[C]//Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001, 2: 416-423.
[19] Arbelaez P, Maire M, Fowlkes C, et al. Contour detection and hierarchical image segmentation[J]. IEEE Trans. Pattern Anal. Mach. Intell., 2011, 33(5): 898-916. DOI: 10.1109/TPAMI.2010.161
[20] Silberman N, HoiemD, Kohli P, et al. Indoor segmentation and support inference from RGBD images[C]//European Conference on Computer Vision, 2012: 746-760.
[21] Mottaghi R, CHEN X, LIU X, et al. The role of context for object detection and semantic segmentation in the wild[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 891-898.
[22] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[J/OL]. arXiv preprint arXiv: 1409.1556, 2014.
[23] LI N, SU Z, CHEN Z, et al. A real-time aircraft infrared imaging simulation platform[J]. Optik International Journal for Light & Electron Optics, 2013, 124(17): 2885-2893. http://or.nsfc.gov.cn/bitstream/00001903-5/94940/1/1000006643792.pdf
[24] Davis J, Keck M. A Two-Stage Template Approach to Person Detection in Thermal Imagery[C]// 2005 Seventh IEEE Workshops on Applications of Computer Vision, 2005, 1: 364-369.
[25] WANG Z, Bovik A C, Sheikh H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Trans Image Process, 2004, 13(4): 967-981. http://citeseer.ist.psu.edu/viewdoc/download;jsessionid=1CEF39B23BCE926AD6D155479B134452?doi=10.1.1.11.2477&rep=rep1&type=pdf
[26] ZHANG L, ZHANG L, MOU X, et al. FSIM: A Feature Similarity Index for Image Quality Assessment[J]. IEEE Transactions on Image Processing, 2011, 20(8): 2378-2386. DOI: 10.1109/TIP.2011.2109730
-
期刊类型引用(1)
1. 肖沁，李正周，刘海毅. 基于场景自适应方向引导滤波的红外成像非均匀性校正方法. 光子学报. 2024(11): 253-265 . 
百度学术
其他类型引用(0)
计量
- 文章访问数:
- HTML全文浏览量:
- PDF下载量:
- 被引次数: 1
