首页> 外文会议>Multispectral, hyperspectral, and ultraspectral remote sensing technology, techniques, and applications II >A method for detecting change in coral reef using pan-sharpened satellite images
【24h】

A method for detecting change in coral reef using pan-sharpened satellite images

机译:一种使用泛锐化卫星图像检测珊瑚礁变化的方法

获取原文
获取原文并翻译 | 示例

摘要

A method was developed for global monitoring of temporal change of coral reef using pan-sharpened color images with higher accuracy and lower cost. The method consisted of 3 blocks; image co-registration for removing complex discrepancy due to parallax among original color image and panchromatic one, pan-sharpening with preserving color information, and change detection with suppressing noise such as sea waves. The method was successfully applied to an actual FORMOSAT2 multi-temporal data set. After removing the parallax between multi-spectral band images and panchromatic one, the spatial resolution of multi-spectral images was improved from 8 × 8 m~2 to 2 × 2 m~2 in the following pan-sharpening block. The pan-sharpening was performed by replacing brightness component of the original multi-spectral pixel with the panchromatic pixel density. In order to normalize the recording gain and offset, the brightness component was estimated by using a linear polynomial model whose coefficients were determined by applying a multiple regression analysis. Linear shapes of density scatter diagram of each spectral band between pan-sharpened density and original one indicated that the pan-sharpened spectral information was perfectly preserved. The change detection successfully detected some temporal changes with suppressing noise. The method was applicable to other data sets having lower resolution multi-spectral images and panchromatic one covering all spectral bands.
机译:开发了一种方法,该方法使用泛锐化的彩色图像以更高的准确性和更低的成本来全局监视珊瑚礁的时间变化。该方法由3个块组成;图像共配准,用于消除原始彩色图像和全色图像之间的视差引起的复杂差异,保留色彩信息的全色锐化以及抑制海浪等噪声的变化检测。该方法已成功应用于实际的FORMOSAT2多时间数据集。消除多光谱带图像和全色图像之间的视差后,在随后的全锐化块中,多光谱图像的空间分辨率从8×8 m〜2提高到2×2 m〜2。通过将原始多光谱像素的亮度分量替换为全色像素密度来执行全色锐化。为了归一化记录增益和偏移,通过使用线性多项式模型来估计亮度分量,该线性多项式模型的系数通过应用多元回归分析来确定。泛锐化密度和原始锐化之间的每个光谱带的密度散布图的线性形状表明,泛锐化光谱信息得到了完美的保留。变化检测在抑制噪声的情况下成功地检测到一些时间变化。该方法适用于具有较低分辨率的多光谱图像和覆盖所有光谱带的全色图像的其他数据集。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号