A new way for the detection and recognition of bridges

In this paper, a two-step scheme based on the maximum variance cluster algorithm was presented to detect and recognize bridges over water with the complicated background from the remote sensing images, First, the water including bridges was detected in all region of image; then, image segmentation and feature extraction and recognition were performed over the water having been detected. According to the above thought, this paper presented one miniaturized system that is realized by hardware, which is based on the DSP and possess low power loss. And the remote sensing images about bridge are done some experiment, the result of experiment indicated this method for detecting and recognizing the bridge over waters is feasible and effective.