Research for Target Recognition of Infrared Bridge Based on Morphological Operator and Bridge Template

A novel algorithm of infrared bridge target recognition, based on bridge template and bridge-knowledge, was presented. Firstly, some operations including contrast stretching transformation, Otsu operator, morphological operators were employed to process the original image. Secondly, bridge template of four directions is defined and some rules for finding possible pixel points on bridge is developed, Thirdly, the skelton processing and hough tranformation is used to find a possible bridge target. Finally, some infrared images including bridge targets were used to make simulation experiment, and the results show that the novel algorithm has better performance of accuracy and reliability than traditional algorithm, it has lower error and higher computational efficiency and can be used to real-time application.