含明亮区域的无人机遥感定位图像去雾方法  

Dehazing Method for the Remote Sensing and Positioning Image of UAV with Bright Area

作　　者：黄莺[1] 胡凯益 李战一 黄鹤[2,3] 茹锋 HUANG Ying;HU Kaiyi;LI Zhanyi;HUANG He;RU Feng(Air Force Engineering University,Xi’an 710038,China;Chang’an University,Xi’an 710064,China;Xi’an Key Laboratory of Intelligent Expressway Information Fusion and Control,Xi’an 710064,China)

机构地区：[1]空军工程大学,西安710038 [2]长安大学,西安710064 [3]西安市智慧高速公路信息融合与控制重点实验室,西安710064

出　　处：《火力与指挥控制》2024年第5期130-136,144,共8页Fire Control & Command Control

基　　金：国家自然科学基金面上项目(52172379);陕西省重点研发计划项目(2021SF-483);西安市智慧高速公路信息融合与控制重点实验室(长安大学)开放基金资助项目(300102323502)。

摘　　要：针对传统DCP去雾算法处理无人机遥感定位含雾图像时,天空或白色等明亮区域颜色易发生失真,图像整体对比度降低等问题,提出了一种自适应阈值分割的DCP去雾方法。利用灰度图像I_(gray)(x)求取图像明亮与非明亮区域的自适应阈值ThrB;根据自适应阈值ThrB将明亮区与非明亮区分割,并设计自适应修正函数M;优化由暗通道图像生成的大气耗散函数粗估计,利用双边滤波再次细化透射率,完成图像去雾复原。实验结果表明:提出方法在处理天空或反光较强的明亮区域时,能够有效避免复原后的颜色失真等问题,进一步改善遥感图像地面景物区域的处理效果,复原后整幅遥感图像的色彩饱和度和对比度明显提高,主观视觉效果有一定改善,且PSNR、FC、SSIM和CR等客观参数均有提升,有利于后续遥感定位图像分析。As the traditional dark channel prior(DCP)dehazing algorithm processes the remote sensing positioning image with hazeof UAV,the color of such bright areas as the sky or white is prone to distortion,the overall contrast of the image is reduced,and the areas of close-up scenery in the image are prone to such problems as noise patches and blurred details after processing.An adaptive threshold segmentation dark channel prior dehzaing method is proposed.Firstly,the grayscale image I_(gray)(x)is used to obtain the adaptive threshold ThrB of the bright and non-bright areas of the image.Then,according to the adaptive threshold ThrB,the bright and non-bright areas on the image can be segmented,and the adaptive correction function M is designed for these regions.Finally,the coarse estimate of the atmospheric dissipation function generated by the dark channel image is optimized,and the transmittance is refined again with the bilateral filtering to complete the image dehazing restoration.The experimental results show that the proposed method can effectively avoid the problem of color distortion after image restoration when the sky area or the bright area with strong reflection is processed,and can further improve the processing effects of the ground area in the remote sensing image.After restoration,the color saturation and contrast of the entire remote sensing image are obviously improved,the subjective visual effect is improved to a certain extent,and the objective parameters such as PSNR,FC,SSIM,and CR are all improved,which is conducive to the subsequent remote sensing positioning image analysis.

关 键 词：图像处理 暗通道理论 去雾 遥感 定位 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]