逆光海况下低质量红外目标的增强与识别  

Enhancement and Recognition of Infrared Target with Low Quality Under Backlight Maritime Condition

作　　者：侯义锋 丁畅 刘海 Mandal Mrinal 高兴宇[2] 罗祯东 吴自库[5] Hou Yifeng;Ding Chang;Liu Hai;Mandal Mrinal;Gao Xingyu;Luo Zhendong;Wu Ziku(School of Electronics and Information,Wuzhou University,Wuzhou Guangxi,543001,China;School of Mechanical and Electrical Engineering,Guilin University of Electronic Technology,Guilin Guangxi,531004,China;School of Information and Control Engineering,China University of Mining and Technology,Xuzhou,Jiangsu,221116,China;Department of Electrical and Computer Engineering,University of Alberta,Edmonton,Alberta,T6G 1H9,Canada;School of Science and Information,Qingdao Agricultural University,Qingdao,266109,China)

机构地区：[1]梧州学院电子与信息工程学院,广西梧州543001 [2]桂林电子科技大学机电工程学院,广西桂林531004 [3]中国矿业大学信息与控制工程学院,江苏徐州221116 [4]阿尔伯塔大学电子与计算机工程系,埃德蒙顿阿尔伯塔加拿大T6G 1H9 [5]青岛农业大学理学与信息科学学院,山东青岛266109

出　　处：《光学学报》2023年第6期218-230,共13页Acta Optica Sinica

基　　金：国家自然科学基金(62162054);中国国家留学基金(202108455015);广西自然科学基金(2020GXNSFBA297077);桂林市科学研究与技术开发计划(20210217-17);认知无线电与信息处理省部共建重点实验室基金(CRKL200103);梧州学院校级科研重点项目(2022B010)。

摘　　要：海上红外目标识别在海事搜寻中发挥着重要的作用,针对逆光海况下出现的目标反向对比度特点及其目标淹没于背景的问题,结合其直方图呈现的局部尖峰特点,提出了修正灰度占比的新直方图均衡化并融合边缘信息的增强算法。该算法可以有效地提升目标区域的对比度,从而提升海上红外图像的质量。在海上红外目标识别与检测中,建立了目标与背景的多尺度“九宫格”搜索框,演化了局部对比度显著性量化的数学模型,实现了符合人眼视觉特性的红外目标的准确定位与检测。在海上红外图像增强测试中,所提算法可以使原图像的平均梯度提升两倍以上,使局部对比度增益因子提升两倍以上。在目标识别的算法测试中,所提多尺度局部对比度目标检测算法可以使目标检测率达到99%以上。Objective Infrared maritime target recognition plays a significant role in the field of maritime search and maritime rescue.However,complex maritime condition and illumination interference decrease the maritime infrared image’s quality,for example,backlight maritime condition can result in maritime target with negative contrast and disappeared in the maritime background.How to enhance the maritime infrared target’s contrast and improve the dim target’s quality in the backlight maritime condition is the significant basis of maritime target’s detection and recognition,furthermore,studying the effective method on the detection of maritime target in the backlight maritime condition is a significant research direction,and conventional target detection’s algorithm often leads to high false alarm rate and low detection rate.We hope that our proposed enhancement algorithm can outstand the maritime target in the backlight maritime condition and our proposed target detection’s algorithm can detect the maritime target with high accuracy rate in the backlight maritime condition.Methods We test the histograms of multiple maritime infrared images in the backlight condition and studied the characteristic of histograms,then we found that the histograms reflect the characteristic of local peak which represents the large background in the backlight maritime condition.We limit the maximum proportion in the histogram and obtain the novel histogram,then the novel histogram is equalized which can prevent the illumination interference and improve the image’s quality,then we quantize and extract the edge information in the backlight maritime image,finally the edge information is fused with suitable proportional parameter in the middle result of novel histogram equalization,which make the enhancement result with high contrast and more details information.Conventional method for target detection such as Local Contrast Method(LCM)is used in the target detection with positive contrast by target detection unit with single scale

关 键 词：测量 红外成像 海上红外目标识别 红外图像增强 直方图变换 局部对比度显著性 逆光海况 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]