资源简介

摘要： 遥感图像条带噪声掩盖数据真实的辐射信息，降低图像的质量及应用效果。针对窗口矩匹配算法对灰度分布不均一影像的条带噪声去除不彻底问题,提出小波矩匹配条带噪声去除算法。首先，基于小波多分辨率的特性将低频小波系数和高频小波系数分离；其次，利用移动窗口矩匹配算法对影像低频成分滤波，利用阈值法去除影像高频成分中的条带噪声；最后，通过小波逆变换重构得到去条带后的影像。采用局部峰值信噪比、局部结构相似度、模糊系数和拟合优度对影像去噪效果进行定量评价。结果表明:小波矩匹配算法的各项指标均优于矩匹配、小波软阈值和窗口矩匹配算法。小波矩匹配算法改善了窗口矩匹配算法要求影像灰度分布均匀的局限性，结合空间域和频率域去噪的优势，在去除条带噪声的同时能够较好地保留影像的细节信息。 Abstract: Strip noise will cover the real radiation information of remote sensing image，which not only reduces image quality，but also application effect.Aiming at the problem that window moment matching is difficult to remove strip noise thoroughly for the non\|uniform image，the wavelet moment matching algorithm is proposed.Firstly，the low frequency wavelet coefficient and the high frequency wavelet coefficients are separated which based on the wavelet multiresolution characteristic.Secondly，the low frequency wavelet coefficients were filtering by window moment matching，while strip noise of high frequency wavelet coefficients was performed by threshold method.Finally，reconstruction of the wavelet coefficients obtained destriping image.The above mentioned algorithm was evaluated quantitatively by local PSNR，local SSIM，Fuzzy coefficient and goodness of fit index.Results show that the wavelet moments matching algorithm is superior to the moment matching，wavelet soft threshold and window moment matching algorithm.Wavelet moment matching improves limitation of uniform distribution of grayscale value based on window moment matching，combining the advantages of spatial and frequency domain denosing，which can remove the noise while preserving the details of the image.

相关资源

1. 水质遥感监测的关键要素叶绿素a的反演算法研究进展
2. 基于无人机遥感技术的森林参数获取研究进展
3. 基于无线传感器网络的森林生态系统观测试验平台构建
4. 基于无人机激光雷达的森林冠层高度分析
5. 多源时序国产卫星影像的森林火灾动态监测
6. 基于无人机遥感技术的台风灾害倒伏绿化树木检测

下载说明

·如果您发现该资源不能下载，请通知管理员.gissky@gmail.com

·为确保下载的资源能正常使用，请使用[WinRAR v3.8]或以上版本解压本站资源,缺省解压密码www.gissky.net ，如果是压缩文件为分卷多文件，请依次下载每一个文件，并按照顺序命名为1.rar,2.rar,3.rar...，然后鼠标右击1.rar解压.

·为了保证您快速的下载速度，我们推荐您使用[网际快车]等专业工具下载.

·站内提供的资源纯属学习交流之用,如侵犯您的版权请与我们联系.