Abstract: This paper analyzes the producing process of shadows in remote sensed imagery from theoretical perspective，and derives the quantitative relationship between DN values of the same pixel respectively non-shadowed and shadowed.By comparing restoration result of shadows in the same QuickBird imagery by means of methods such as Histogram Matching，linear-correlation correction，and Gamma correction，choose the suited method.The smoothness of imagery is improved through the handling of edge areas of shadows and technology of image smoothing.Experiment shows that the proposed method is effective.

摘要： 从理论上研究了遥感图像阴影产生的过程，得到了同一象素点在无阴影与有阴影时灰度值的定量关系。将分别采用直方图匹配法、线性相关修正法和Gamma 修正法得到的同一QuickBird影像阴影区图像重建结果进行对比，选择影像重建最适合的方法。再通过阴影边界区处理和图像平滑技术，提高图像平滑过渡特性。实验结果表明了该方法的有效性。