[1]沈 旭,程小辉,王新政.结合视觉注意力机制基于尺度自适应局部对比度增强的 红外弱小目标检测算法 [J].红外技术,2019,41(8):764-771.[doi:10.11846/j.issn.1001_8891.2019080012]
 SHEN Xu,CHENG Xiaohu,WANG Xinzheng. Infrared Dim-small Object Detection Algorithm Based on Adaptive Scale Local Contrast Enhancement Combined with Visual Attention Mechanism[J].Infrared Technology,2019,41(8):764-771.[doi:10.11846/j.issn.1001_8891.2019080012]
点击复制

结合视觉注意力机制基于尺度自适应局部对比度增强的
红外弱小目标检测算法
分享到:

《红外技术》[ISSN:1001-8891/CN:CN 53-1053/TN]

卷:
41卷
期数:
2019年第8期
页码:
764-771
栏目:
出版日期:
2019-08-21

文章信息/Info

Title:
 Infrared Dim-small Object Detection Algorithm Based on Adaptive Scale Local Contrast Enhancement Combined with Visual Attention Mechanism
文章编号:
1001-8891(2019)08-0764-08
作者:
沈 旭1程小辉2王新政2
1. 岭南师范学院 信息工程学院,广东 湛江 524048;2. 桂林理工大学 信息科学与工程学院,广西 桂林 541004
Author(s):
SHEN Xu1CHENG Xiaohui2WANG Xinzheng2
1. School of Information Engineering, Lingnan Normal University, Zhanjiang 524048, China;
2. College of Information Science and Engineering, Guilin University of Technology, Guilin 541004, China
关键词:
目标检测视觉注意尺度感知对比度测量跳出效应
Keywords:
object detection visual attention scale perception contrast measurement pop-out effect
分类号:
TN219
DOI:
10.11846/j.issn.1001_8891.2019080012
文献标志码:
A
摘要:
 如何在没有先验信息的情况下从复杂噪声背景下快速检测到远距离进入的弱小目标,提高整个装备系统的响应能力,是目前IRST热门研究课题。本文通过引入视觉注意机制,提出了一种结合尺度自适应的局部对比度测量的红外弱小目标检测方法。本文首先采用拉普拉斯金字塔尺度空间理论对所有像素点局部对比度进行分析,获得对应的自适应尺度信息;然后在跳出效应的基础上设计了一种基于改进的局部对比度测量模型,最终生成一个显著图来突出目标特性,该方法能够在增强目标对比度同时,抑制背景杂波。定性定量实验结果表明,本文提出的方法相比于对比算法具有较高的红外小目标检测性能,能够对对比度不低于5%的目标稳定检测,适合防空武器装备工程应用。
Abstract:
Infrared search and track (IRST) is one of the indispensable key systems in the field of modern warfare. It is widely used in precision guidance, photoelectric warning, air defense, and anti-missile. Methods for quick detection of dim small object entering from a complex background without a priori information and improvement of the responsiveness of the entire equipment system is currently a trending research topic in IRST. In this paper, the visual attention mechanism is introduced, and an infrared dim small object detection method based on local scale-perception contrast measurement is proposed. First, the local contrast of all the pixels is analyzed using Laplacian pyramid scale-space theory, wherein the corresponding adaptive scale information is obtained. Next, according to the pop-out effect in visual system, an improved local contrast measurement model is designed to obtain the optimal saliency map. Finally, a weight map is generated to highlight the object features, thereby suppressing the background clutter while enhancing the object contrast. Qualitative and quantitative experiments show that our proposed method has better performance in terms of superior object detection than the algorithms it was compared with and can detect object with better stability and a contrast of at least 5%, thus ensuring the method’s suitability for engineering application.

参考文献/References:

[1] 杨元庆, 张志利, 侯传勋. 一种近地背景下红外弱小目标检测预处理算法[J]. 红外技术, 2018, 40(8): 812-817.
YANG Yuanqing, ZHANG Zhili, HOU Chuanxun. A preprocessing algorithm for infrared small-target detection in the near-earth background[J]. Infrared Technology, 2018, 40(8): 812-817.
[2] 史漫丽, 凌龙, 吴南, 等. 基于自适应Gabor滤波的红外弱小目标检测[J]. 红外技术, 2018, 40(7): 632-637.
SHI Manli, LING Long, WU Nan, et al. Dim infrared target detection based on adaptive Gbor filter[J]. Infrared Technology, 2018, 40(7): 632-637.
[3] 杨慧, 张宝辉, 沙涛, 等. 地空背景下红外弱小运动目标检测[J]. 红外技术, 2018, 40(5): 462-467.
YANG Hui, ZHANG Baohui, SHA Tao, et al. Detection of small infrared moving targets under ground-sky background[J]. Infrared Technology, 2018, 40(5): 462-467.
[4] 田广强. 一种新颖高效的红外动态场景多目标检测跟踪[J]. 红外技术, 2018, 40(3): 259-263.
TIAN Guangqiang. A novel algorithm for efficient multi-object detection and tracking for infrared dynamic frames[J]. Infrared Technology, 2018, 40(3): 259-263.
[5] Deshpande S D, Meng H E, Venkateswarlu et al. Max-mean and max-median filters for detection of small targets[C]// Proceedings of SPIE, 1999, 3809: 74-83.
[6] WANG Qi, LIU Gang, SHI Yaovw. Detecting of multi-dim- small-target in sea or sky background based on higher-order cumulants and wavelet[J]. Recent Advances in Computer Science and Information Engineering, 2012: 497-504.
[7] HAN J H, LIANG K, ZHOU B, et al. Infrared small target detection utilizing the multiscale relative local contrast measure[J]. IEEE Geosci. Remote Sens. Lett., 2018, 15(4): 612-616.
[8] CUI Z, YANG J L, LI J B, et al. An infrared small target detection framework based on local contrast method[J]. Measurement, 2016, 91: 405-413.
[9] WEI Y T, YOU X G, LI H, et al. Multiscale patch-based contrast measure for small infrared target detection[J]. Pattern Recognit., 2016, 58: 216-226.
[10] BAI X , BI Y . Derivative entropy-based contrast measure for infrared small-target detection[J]. IEEE Transactions on Geoscience & Remote Sensing, 2018, 99: 1-15.
[11] DENG H , SUN X , ZHOU X. A multiscale fuzzy metric for detecting small infrared targets against chaotic cloudy/sea-sky backgrounds[J]. IEEE Transactions on Cybernetics, 2018, 99: 1-14.
[12] 康红霞, 黄树彩, 黄文龙, 等. 最小二乘向量机结合光谱维信息检测弱小目标[J]. 现代防御技术, 2018, 46(4): 86-91, 119.
KANG H, HUANG S et al. The least squares vector machine combined with spectral dimension information to detect weak targets[J]. Modern Defence Technology, 2018, 46(4): 86-91, 119.
[13] 荣楚君, 曹晓光, 白相志. Facet方向导数特征与稀疏表示相结合的红外弱小目标检测算法[J]. 中国图象图形学报, 2018, 23(11): 1768-1776.
RONG C, CAO X, BAI Xiangzhi. Infrared weak small target detection algorithm based on fisher’s derivative features and sparse representation[J]. Journal of Image and Graphics, 2018, 23(11): 1768-1776.
[14] ZHOU Y, ZHANG J, LIN X. Infrared small target detection using weighting LoG operator[J]. Journal of Applied Optics, 2017, 38(1): 114-119.
[15] HE D, SUN X, LIU M, et al. Entropy-based window selection for detecting dim and small infrared targets[J]. Pattern Recognition, 2017, 61: 66-77.

相似文献/References:

[1]张双垒,林剑春,段东,等.基于遗传算法红外小目标检测的研究[J].红外技术,2012,34(08):472.
 ZHANG Shuang-lei,LIN Jian-chun,DUAN Dong,et al.Infrared Small Target Detection Research Based on Genetic Algorithm[J].Infrared Technology,2012,34(8):472.
[2]李旭,赵文杰,杨凯达.基于小目标预提取的OTSU分割方法[J].红外技术,2013,35(08):492.[doi:10.11846/j.issn.1001_8891.201308008]
 LI Xu,ZHAO Wen-jie,YANG Kai-da.OTSU Applied in Image Segmentation Based on Small Targets Pre-detection[J].Infrared Technology,2013,35(8):492.[doi:10.11846/j.issn.1001_8891.201308008]
[3]王宇翔,韩振铎,王宏敏.基于多向差异度的红外弱小目标检测算法[J].红外技术,2012,34(06):351.
 WANG Yu-xiang,HAN Zhen-duo,WANG Hong-min.Detection Algorithm for Dim Infrared Target Based on Multi-Difference Factor[J].Infrared Technology,2012,34(8):351.
[4]李国强.高斯曲面拟合在对空红外点目标检测中的应用[J].红外技术,2013,35(10):638.[doi:10.11846/j.issn.1001_8891.201310008]
 LI Guo-qiang.Application of Gaussian Quadrics Fitting in the Infrared Point Targets Detection in Sky Background [J].Infrared Technology,2013,35(8):638.[doi:10.11846/j.issn.1001_8891.201310008]
[5]曾雅琼,陈钱.基于改进的双边滤波的单帧红外弱小目标背景抑制[J].红外技术,2011,33(09):537.
 ZENG Ya-qiong,CHEN Qian.Dim and Small Target Background Suppression?Based on Improved Bilateral Filtering for Single Infrared Image[J].Infrared Technology,2011,33(8):537.
[6]何永强,周云川,仝红玉,等.基于YCBCR空间颜色传递的融合图像目标检测算法[J].红外技术,2011,33(06):349.
 HE Yong-qiang,ZHOU Yun-chuan,TONG Hong-yu,et al.A Target Detection Method Based on Color Transferin YCBCR Space for Fused Image[J].Infrared Technology,2011,33(8):349.
[7]王莹莹,张永顺,华永伟.红外目标检测方法分析[J].红外技术,2011,33(03):133.
 WANG Ying-ying,ZHANG Yong-shun,HUA Yong-wei.Analysis on Infrared Target Detection Methods[J].Infrared Technology,2011,33(8):133.
[8]邹志峰,张艳红,张健.基于海空背景的红外图像处理算法研究[J].红外技术,2011,33(02):108.
 ZOU Zhi-feng,ZHANG Yan-hong,ZHANG Jian.Theoretical Methods of Infrared Image Signal Processingunder Sea and Sky Background[J].Infrared Technology,2011,33(8):108.
[9]郭明,王学伟.基于红外传感器参数的红外/主动雷达弱小目标融合检测方法[J].红外技术,2010,32(8):475.
 GUO Ming,WANG Xue-wei.An IR/Active Radar Fused Target Detection MethodBased on IR Sensor Parameters[J].Infrared Technology,2010,32(8):475.
[10]王斯博,杨春玲.基于RX算法的高光谱红外弱小目标检测[J].红外技术,2010,32(4):204.
 WANG Si-bo,YANG Chun-ling.Detection of Dim Small Target Based on RX algorithmfor Hyperspectral IR Imagery[J].Infrared Technology,2010,32(8):204.

备注/Memo

备注/Memo:
收稿日期:2018-12-28;修订日期:2019-04-08.
作者简介:沈旭(1979-),男,硕士,讲师,主要研究方向图像处理、模式识别、智能控制应用等。
基金项目:国家自然科学基金(61662017,61402399);广东省哲学社会科学规划项目(GD17XGL33);岭南师范学院教育教学改革项目(LSJGMS1811)。
更新日期/Last Update: 2019-08-20