A recognition algorithm based on knowledge framework for long-range infrared bridge images

Through the analysis of the features of bridge object in long-range IR image, a new recognition algorithm based on knowledge framework is put forward. Due to some characters of long-range infrared bridge images, the global method processing the whole image equally is ineffective. Based on the features of the bridge object, our proposed algorithm first detects the candidate bridge regions and then focuses on them. According to the degree to which they match our prebuilding framework, different credits are given, so the false targets are excluded and eventually the real target is found. The experiment results demonstrate the good performance of our method.