Tank target recognition used in infrared imaging fuze based on FPGA

Infrared imaging fuze is invulnerable to the electromagnetic interference, and it has the ability to recognize the local image of the target. At present, the infrared imaging fuze technology has become one of the key technologies which perform the target detection and the ignition of the warhead in the complex tactical environment. According to the scanning mechanism of the infrared imaging fuze, based on the analysis of features of the infrared image of tank target, this paper presents a feature extraction method based on knowledge to recognize infrared gray image. The geometric features and gray level features are extracted. The geometric features include the corner features and angular features. The corners of the image are extracted through the SUSAN corner detection principle,the angular feature is extracted by Freeman chain code. The hot-zone gray feature is extracted by the template matching and image binarization principle. In order to realize real-time recognition, this paper uses FPGA technology to achieve recognition circuit. The experiments show that the recognition method has a certain anti-interference ability.