针对弹载平台红外小目标的检测与跟踪算法

为了提高对红外小目标检测与跟踪的精度与速度，设计了一种针对红外小目标的单目标检测与跟踪算法,包含一个基于MB_LBP+AdaBoost与管道滤波器的目标检测模块，和一个有跟踪失败监测机制的基于模板匹配的多尺度跟踪模块。经过测试，该算法对边长11~31像素大小的指定类型的红外小目标取得了较好的检测效果，目标跟踪模块对于目标的尺度变化，快速运动，遮挡有较高的鲁棒性，算法中的跟踪失败监测机制在检测到跟踪异常的情况下能重新调用目标检测算法找回目标，经过测试对比，本方法跟踪的精度和速度综合优于主流跟踪算法，并且能够在弹载平台上实时运行。

Detection and Tracking Algorithm for Small Infrared Target on Missile Borne Platform

In order to improve the performance of small infrared target detection and tracking, a single target detection and tracking algorithm is proposed, in which the detection module is based on the MB_LBP+AdaBoost algorithm and pipeline filter, and the tracking module with multi scale search and tracking failure monitoring mechanism is based on template matching. After testing, the proposed algorithm has good detecting performance for the small infrared targets of the specified type with a side length of 11~31 pixels. The tracking module has high robustness to target scale changes, fast motion and occlusion. The tracking failure monitoring mechanism can recall the detection module to retrieve the target after tracking failure. After comparison, the tracking performance is better than the mainstream tracking algorithms, and can perform real time running on the missile platform.