The automatic recognition of bridges has both civil and military significance. However, in complicated cases when the image resolution is at the decimeter scale. the bridge scenes are messy and the targets small, and automatic recognition will become quite complicated. Thus, we proposed a novel algorithm based on the analysis of the statistical distribution and features of bridge targets in high-resolution SAR images. A CFAR detector locates potential bridge targets based on the Weibull distribution. Scene areas of bridges are extracted and false alarms are removed by utilizing the features of bridges with the help of Hough transformation. Domestic airborne polarimetric SAR data and AIRSAR data illustrate the effectiveness of this method. Results indicate that this algorithm recognizes bridges in complicated cases with high adaptability.