Man-made target detection for imaging infrared terminal guidance missile considering turbulence degraded Wave-ronts features

Previous research employed natural background fractal features to detect man-made target for imaging infrared terminal guidance missile. The method is some effective but the optical turbulence is overlooked, which will disturb natural background intrinsic fractal. When the missile flies in atmosphere, there exits turbulent flow over the IR windows which will degrade the optical wave-fronts. In this paper, a new method considering degraded wave-fronts features is proposed for man-made moving target detection in natural background. To pre-process the infrared image and obtain the area in which target may exit, the image is divided into blocks and each block fractal dimension is calculated and compared. Then optical flow of the block is calculated from successive images to determine the moving target. The method is more applicable to actual missile fly environment, simulation results show that it reduces the optical flow calculation complexity and can detect the target availably.