如何自动、正确且快速地从遥感或航拍图片中检测战略目标 ,是导弹能否准确击中目标的关键 ,在军事领域具有重要意义。提出了一种从航片中对桥梁自动检测的算法。该方法从河流的识别入手 ,利用小波变换的阈值分割法和高斯 -拉普拉斯变换的分割法 ,将河流部分提取出来并二值化。然后对其进行多步噪声滤除从而使河流的提取更加完整。最后沿着河流搜索有无桥梁存在。实验结果表明 ,对于航空图片中存在的桥梁 ,该算法均能正确地加以检测 How to automatically, correctly and quickly detect strategic targets from remote sensing or aerial photographs is the key to whether missiles can accurately hit targets and is of great significance in the military field. An algorithm for automatic bridge detection from aerial photos is proposed. The method starts with the identification of rivers, and uses the threshold segmentation method of wavelet transform and the segmentation method of Gauss - Laplace transform to extract and binarize the river parts. Then it is multi-step noise filtered to make the extraction of the river more complete. Finally there is a bridge search along the river. The experimental results show that this algorithm can correctly detect the bridges existing in the aerial picture