多目标航迹的图像域自适应检测方法

摘要

Aiming at the problem of low visible moving target track detection, combining image processing and curve fitting, an approach of multi-target detection is presented. On the basis of range-time image, which is accessed through processing radar echo, using phase congruency model to detect edge, the edge connected region with congruency edge orientation is obtained by combining adaptive region growing technology. The target moving track and moving parameters are estimated using clearing technology and least squares fitting. The simulation results show that, the method can detect multi-target track under low signal noise ratio, and it is difficult to detect multi-target track with phase congruency model.%针对低可观测运动目标航迹检测问题，提出了一种融合图像处理和曲线拟合的多目标航迹检测方法。在对雷达回波进行处理获得距离-时间图像的基础上，利用相位一致性模型检测边缘，结合自适应区域生长技术提取具有一致性边缘走向的连通区域；采用clearing技术和迭代最小二乘方法估计目标运动轨迹和运动参数。仿真结果表明：该算法即使在信噪比很低时，也能清晰地检测到多目标航迹，而仅使用相位一致性模型则很难检测出多目标航迹。