摘要
森林雾天常导致采集的图像信息质量差,为得到更为清晰完整的森林去雾图像,从而为森林监测提供更好的数据支持和保障,该文利用自适应性算法初步确定森林雾气含量以及光线强度等环境状况,再用亮暗通道融合算法对全局大气光、透射率等大气模型的重要参数值进行优化,最后针对亮暗通道先验算法造成的晕轮效应进行高斯曲率滤波处理。基于自适应图像增强的亮暗通道融合去雾算法能有效改善森林图像的质量,得到细节丰富、视觉效果清晰的森林去雾图像。实验分别从主观标准和客观标准对图像效果进行科学性评价,所得到的融合改进算法对森林图像优化有良好的效果。
Forest fog often leads to poor quality of collceted image information.In order to obtain clearer and complete images of forest defogging images and provide better data support and guarantee for forest monitoring,adaptive algorithm is used to preliminarily determine the forest fog content,light intensity and other environmental conditions.Then,the light and dark channel fusion algorithm is used to optimize the global atmospheric light,transimittance and other important parameters of the atmospheric model.Finally,Gaussian curvature filtering is used to deal with the halo effect caused by the light and dark channel prior algorithm.Light and dark channel fusion defogging algorithm based on adaptive image enhancement can effectively improve the quality of forest images,and obtain clear and complete forest defogging images with rich details and clear visual effects.The experiment evaluates the image effect scientifically from subjective standard and objective standard respetively,and the fusion improved algorithm obtained has good effect on forest image optimization.
作者
孟宇彤
赵康军
赵伟
MENG Yutong;ZHAO Kangjun;ZHAO Wei(College of Information and Computer Engineering,Northeast Forestry University,Harbin 150040,China;Linshu Industrial and Information Industry Development Center,Linyi 276700,China)
出处
《森林工程》
北大核心
2022年第4期106-112,共7页
Forest Engineering
基金
国家自然科学基金项目(61975028)
黑龙江省重点研发计划(GZ20210017)
黑龙江省重点研发计划(GZ20210018)。
关键词
森林图像优化
自适应
图像去雾
亮暗通道融合
高斯曲率滤波
Forest image optimization
self-adaption
image defogging
light and dark channel fusion
Gaussian curvature filtering
