改进的基于图像边缘的能见度反演方法  

Improved inversion algorithm based on image edge for visibility

作　　者：吕艳芬[1,2] 杨玲[1,2] 甄小琼[1] 

机构地区：[1]成都信息工程学院电子工程学院,成都610225 [2]中国气象局大气探测重点开放实验室,成都610225

出　　处：《计算机工程与应用》2015年第10期170-173,256,共5页Computer Engineering and Applications

基　　金：公益性行业(气象)科研专项课题(No.GYHY201106047)

摘　　要：针对现有的基于图像边缘的能见度计算方法效果不理想的问题,提出一种改进的基于图像边缘检测求能见度的方法。相对于原有的方法,该方法首先选取目标图像并用sift算法对其进行模板匹配,然后利用离散Shearlet变换提取边缘特征,再将这些特征和前向散射式能见度仪测量的数据进行线性拟合,建立它们之间的关系式,从而求出能见度。实验结果表明,该算法相对于sobel算法和小波变换算法,相对误差减少了2.761%和1.607%,计算的准确性和有效性明显得到改善。According to the existing method for image edge visibility, the result is not ideal, this paper presents an improved method based on image edge detection for visibility. Firstly, it selects the target image using sift algorithm for template matching, then the discrete shearlet transform is used to extract the edge features, and then these features and the results of forward-scattering visibility meter are related. After establishing the relationship, visibility can be computed by the features with the built relationship. The experimental results show that the relative error is reduced by 2.761% relative to Sobel algorithm, and by 1.607%relative to wavelet transform algorithm. The accuracy and efficiency are obviously improved.

关 键 词：能见度 SHEARLET变换 图像边缘 sift图像匹配 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]