融合视觉显著性和局部熵的红外弱小目标检测
doi: 10.37188/CO.2021-0170
Infrared dim small target detection based on visual saliency and local entropy
-
摘要: 针对红外图像中弱小目标检测虚警率高、实时性差的问题,提出了一种基于视觉显著性和局部熵的红外弱小目标检测方法。该方法将红外弱小目标的检测问题由粗到精分步实现,首先利用融合局部熵的方法提取包含目标的感兴趣区域,对红外弱小目标实现粗定位。然后再利用改进的视觉显著性检测方法在感兴趣区域计算局部对比度,获得感兴趣区域的显著图。最后利用阈值法分割显著图像提取红外弱小目标,实现红外弱小目标的检测。通过与TOPHAT算法及LCM算法进行对比试验,验证了该方法在检测性能上优于TOPHAT算法以及LCM算法,虚警率分别下降了62.5%和33.3%;检测实时性方面,算法耗时为LCM的38.6%。该方法能够实现复杂背景下红外弱小目标的准确检测,在一定程度上解决了目标检测虚警率高、实时性差的问题。
-
关键词:
- 视觉显著性 /
- 红外图像 /
- 弱小目标检测 /
- 局部熵
Abstract: To improve the high false-alarm rate and poor real-time capability in detecting infrared small dim targets, a novel algorithm based on visual saliency and local entropy is proposed in this paper. This method solves the problem from coarse to fine detecting of small targets. First, a local entropy method is used to obtain the region of interest. Then, an improved visual saliency method is used to calculate local contrast. Finally, a threshold segmentation method is used to extract dim infrared small targets. The method is verified using a contrast test with TOPHAT and LCM, and the results show that the performance of this method precedes the TOPHAT algorithm and LCM algorithm. The false alarm rate by this method decreases to 62.5% and 33.3% compared with the other two algorithms, and the time cost decrease to 38.6% of that of LCM. The method can achieve accurate detection of infrared dim and small targets in a complicated environment, solving the high false alarm rate and poor real-time capability issues to some extent.-
Key words:
- visual saliency /
- infrared images /
- dim small target detection /
- local entropy
-
图 1 视觉显著性和局部熵结合的弱小目标检测流程
Figure 1. Block diagram of target detection algorithm combining visual salient and local entropy
图 2 本文算法红外弱小目标检测实验结果
Figure 2. IR small target detection results using the proposed algorithm
图 3 3种目标检测算法对比实验结果
Figure 3. Contrast experiment results with three different target detection algorithms
表 1 3种目标检测算法运行时间对比
Table 1. Computational cost comparison among three target detection algorithms
算法 TOPHAT算法 LCM算法 本文算法 平均耗时/s 0.0307 1.4193 0.5481 PHP网站源码宝安SEO按效果付费民治网站seo优化同乐百搜标王盐田百度爱采购宝安网站推广方案石岩百度网站优化大鹏企业网站设计龙华网页制作大鹏网站改版双龙百度竞价坪地营销型网站建设同乐SEO按天计费民治关键词按天计费东莞百搜词包惠州seo坪山网站搜索优化松岗模板推广大运网站推广方案坑梓网站推广福田网站优化双龙网站改版福田网站搜索优化永湖网络推广南联SEO按效果付费塘坑seo网站优化坪山网站优化按天计费横岗seo网站优化福永网站推广工具吉祥百度seo东莞设计公司网站歼20紧急升空逼退外机英媒称团队夜以继日筹划王妃复出草木蔓发 春山在望成都发生巨响 当地回应60岁老人炒菠菜未焯水致肾病恶化男子涉嫌走私被判11年却一天牢没坐劳斯莱斯右转逼停直行车网传落水者说“没让你救”系谣言广东通报13岁男孩性侵女童不予立案贵州小伙回应在美国卖三蹦子火了淀粉肠小王子日销售额涨超10倍有个姐真把千机伞做出来了近3万元金手镯仅含足金十克呼北高速交通事故已致14人死亡杨洋拄拐现身医院国产伟哥去年销售近13亿男子给前妻转账 现任妻子起诉要回新基金只募集到26元还是员工自购男孩疑遭霸凌 家长讨说法被踢出群充个话费竟沦为间接洗钱工具新的一天从800个哈欠开始单亲妈妈陷入热恋 14岁儿子报警#春分立蛋大挑战#中国投资客涌入日本东京买房两大学生合买彩票中奖一人不认账新加坡主帅:唯一目标击败中国队月嫂回应掌掴婴儿是在赶虫子19岁小伙救下5人后溺亡 多方发声清明节放假3天调休1天张家界的山上“长”满了韩国人?开封王婆为何火了主播靠辱骂母亲走红被批捕封号代拍被何赛飞拿着魔杖追着打阿根廷将发行1万与2万面值的纸币库克现身上海为江西彩礼“减负”的“试婚人”因自嘲式简历走红的教授更新简介殡仪馆花卉高于市场价3倍还重复用网友称在豆瓣酱里吃出老鼠头315晚会后胖东来又人满为患了网友建议重庆地铁不准乘客携带菜筐特朗普谈“凯特王妃P图照”罗斯否认插足凯特王妃婚姻青海通报栏杆断裂小学生跌落住进ICU恒大被罚41.75亿到底怎么缴湖南一县政协主席疑涉刑案被控制茶百道就改标签日期致歉王树国3次鞠躬告别西交大师生张立群任西安交通大学校长杨倩无缘巴黎奥运
-
[1] 黄乐弘, 曹立华, 李宁, 等. 深度学习的空间红外弱小目标状态感知方法[J]. 中国光学,2020,13(3):527-536.HUANG L H, CAO L H, LI N, et al. A state perception method for infrared dim and small targets with deep learning[J]. Chinese Optics, 2020, 13(3): 527-536. (in Chinese) [2] 任向阳, 王杰, 马天磊, 等. 红外弱小目标检测技术综述[J]. 郑州大学学报(理学版),2020,52(2):1-21.REN X Y, WANG J, MA T L, et al. Review on infrared dim and small target detection technology[J]. Journal of Zhengzhou University (Natural Science Edition) , 2020, 52(2): 1-21. (in Chinese) [3] 马铭阳, 王德江, 孙翯, 等. 基于稳健主成分分析和多点恒虚警的红外弱小目标检测[J]. 光学学报,2019,39(8):0810001. doi: 10.3788/AOS201939.0810001MA M Y, WANG D J, SUN H, et al. Infrared dim-small target detection based on robust principal component analysis and multi-point constant false alarm[J]. Acta Optica Sinica, 2019, 39(8): 0810001. (in Chinese) doi: 10.3788/AOS201939.0810001 [4] 周苑, 张健民, 林晓. 基于加权LoG算子的红外弱小目标检测方法研究[J]. 应用光学,2017,38(1):114-119.ZHOU Y, ZHANG J M, LIN X. Infrared small target detection using weighting LoG operator[J]. Journal of Applied Optics, 2017, 38(1): 114-119. (in Chinese) [5] 魏然然, 詹伟达, 朱德鹏, 等. 改进多尺度的Retinex红外图像增强[J]. 液晶与显示,2021,36(3):465-474. doi: 10.37188/CJLCD.2020-0109WEI R R, ZHAN W D, ZHU D P, et al. Improved multi-scale Retinex infrared image enhancement[J]. Chinese Journal of Liquid Crystals and Displays, 2021, 36(3): 465-474. (in Chinese) doi: 10.37188/CJLCD.2020-0109 [6] MAO X, DIAO W H. Criterion to evaluate the quality of infrared small target images[J]. Journal of Infrared,Millimeter,and Terahertz Waves, 2009, 30(1): 56-64. doi: 10.1007/s10762-008-9410-5 [7] 王刚, 陈永光, 杨锁昌, 等. 采用图像块对比特性的红外弱小目标检测[J]. 光学 精密工程,2015,23(5):1424-1433. doi: 10.3788/OPE.20152305.1424WANG G, CHEN Y G, YANG S CH, et al. Detection of infrared dim small target based on image patch contrast[J]. Optics and Precision Engineering, 2015, 23(5): 1424-1433. (in Chinese) doi: 10.3788/OPE.20152305.1424 [8] 何耀民, 何华锋, 徐永壮, 等. 基于改进小波变换的海上目标检测[J]. 系统工程与电子技术,2020,42(1):83-89. doi: 10.3969/j.issn.1001-506X.2020.01.12HE Y M, HE H F, XU Y ZH, et al. Marine target detection based on improved wavelet transform[J]. Systems Engineering and Electronics, 2020, 42(1): 83-89. (in Chinese) doi: 10.3969/j.issn.1001-506X.2020.01.12 [9] DIRAMI A, HAMMOUCHE K, DIAF M, et al. Fast multilevel thresholding for image segmentation through a multiphase level set method[J]. Signal Processing, 2013, 93(1): 139-153. doi: 10.1016/j.sigpro.2012.07.010 [10] SUCCARY R, KALMANOVITCH H, SHURNIK Y, et al. Point target detection[J]. Proceedings of SPIE, 2003, 4820: 671-675. doi: 10.1117/12.453556 [11] 强勇, 焦李成, 保铮. 动态规划算法进行弱目标检测的机理研究[J]. 电子与信息学报,2003,25(6):721-727.QIANG Y, JIAO L CH, BAO ZH. Study on mechanism of dynamic programming algorithm for dim target detection[J]. Journal of Electronics and Information Technology, 2003, 25(6): 721-727. (in Chinese) [12] 杨帆, 汪文英, 王茹琪. 基于粒子滤波TBD的高机动目标检测技术[J]. 中国电子科学研究院学报,2018,13(3):279-283. doi: 10.3969/j.issn.1673-5692.2018.03.008YANG F, WANG W Y, WANG R Q. High maneuvering target detection technology based on particle filter TBD[J]. Journal of China Academy of Electronics and Information Technology, 2018, 13(3): 279-283. (in Chinese) doi: 10.3969/j.issn.1673-5692.2018.03.008 [13] SERENCES J T, YANTIS S. Selective visual attention and perceptual coherence[J]. Trends in Cognitive Sciences, 2006, 10(1): 38-45. doi: 10.1016/j.tics.2005.11.008 [14] ROLLS E T, DECO G. Attention in natural scenes: neurophysiological and computational bases[J]. Neural Networks, 2006, 19(9): 1383-1394. doi: 10.1016/j.neunet.2006.08.007 [15] 刘杨帆, 曹立华, 李宁, 等. 基于YOLOv4的空间红外弱目标检测[J]. 液晶与显示,2021,36(4):615-623. doi: 10.37188/CJLCD.2020-0227LIU Y F, CAO L H, LI N, et al. Detection of space infrared weak target based on YOLOv4[J]. Chinese Journal of Liquid Crystals and Displays, 2021, 36(4): 615-623. (in Chinese) doi: 10.37188/CJLCD.2020-0227 [16] 陈诗媛, 廖一鹏, 张进, 等. 结合NSST显著性检测及图割的泡沫红外图像分割[J]. 液晶与显示,2021,36(4):584-595. doi: 10.37188/CJLCD.2020-0234CHEN SH Y, LIAO Y P, ZHANG J, et al. Foam infrared image segmentation combining NSST saliency detection and graph cuts[J]. Chinese Journal of Liquid Crystals and Displays, 2021, 36(4): 584-595. (in Chinese) doi: 10.37188/CJLCD.2020-0234 [17] ITTI L, KOCH C, NIEBUR E. A model of saliency-based visual attention for rapid scene analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(11): 1254-1259. doi: 10.1109/34.730558 [18] CHEN C L P, LI H, WEI Y T, et al. A local contrast method for small infrared target detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(1): 574-581. doi: 10.1109/TGRS.2013.2242477 [19] 黄果, 许黎, 陈庆利, 等. 非局部多尺度分数阶微分图像增强算法研究[J]. 电子与信息学报,2019,41(12):2972-2979. doi: 10.11999/JEIT190032HUANG G, XU L, CHEN Q L, et al. Research on non-local multi-scale fractional differential image enhancement algorithm[J]. Journal of Electronics &Information Technology, 2019, 41(12): 2972-2979. (in Chinese) doi: 10.11999/JEIT190032 -
![WeChat](/fileZGGX/journal/article/zggx/2022/2/WECHAT/wechat_cn_9dab7c66-ec65-4014-8b9e-320107bb6a42.jpg)
![](/fileZGGX/journal/article/zggx/2022/2/CO-2021-0170-1.jpg)
计量
- 文章访问数: 1007
- HTML全文浏览量: 509
- PDF下载量: 206
- 被引次数: 0