Target Localization on Image-Guided Missile

For image-guided missile, it is important to localize the target and its surrounding area using the images taken by seeker. In this paper, we propose a target localization method that combines on-board IMU data and image information. This method does not require equipment such as GPS or optoelectronic platform, nor does it need prior knowledge or templates of the target. Firstly, a selection strategy is proposed to determine keyframes, which are used to solve the pose of the seeker camera and the 3D coordinate of the target in the camera coordinate system. Then nonlinear optimization is used to reduce the reprojection error and obtain the optimal solution of the seeker camera pose and the 3D coordinate of the target. Finally, the scale uncertainty of the seeker monocular camera is solved by combining the on-board IMU data with images to obtain the coordinate of the target in the shooter coordinate system. The proposed method is validated in the flight data of a certain type of image-guided missile, and a high-precision localization result is obtained at a long missile-target distance.