一种新的城市航空图像阴影检测算法

A NEW SHADOW DETECTION ALGORITHM FOR URBAN AERIAL IMAGES

为实现对城市航空图像中阴影的准确检测,提出一种基于阴影区域颜色性质与统计特性的阴影检测算法。算法中针对阴影的颜色性质构造非线性变换函数,结合阈值分割实现对阴影的初步检测。根据阴影区域统计特性,对检测结果加以修正,避免了对颜色性质与阴影较为相近的地物的误检。实验证明该算法能够准确有效地对城市航空图像中的阴影进行检测。

In order to realise accurate detection on shadow of urban aerial images, we put forward a shadow detection algorithm which is based on colour characteristics and statistical property of shadow area. In the algorithm, the nonlinear transformation function is constructed aiming at colour characteristics of shadow, primary detection on the shadow is implemented in combination with the threshold segmentation methods. The detection results will be corrected in accordance with the statistical property of shadow area, which avoids the fault detection on landmarks which have similar colour characteristics to the shadow. Experiment proves that the algorithm can detect the shadow of urban aerial images accurately and effectively.