复杂背景、强噪声环境中的点源目标检测及跟踪方法

介绍了一种在复杂背景、强嗓声环境中检测并跟踪多个点源目标的方法。目标在图象中的位置是未知的；其轨迹是连续的并可以是任意的；图象序列中含复杂的变化背景，并伴有强的随机白噪声；目标为压几个象元的点目标，本文提出的是一种回形窗扫描、相关处理及相关跟踪相结合的算法，能有效地从复杂背景、强噪声环境中检测与跟踪多个点源目标。这种算法能对图象进行分割、并行处理，适合于实时的目标探测及跟踪。

An Approach for Detecting and Tracking Point Objects From High-Noisy and Complex Background Images

A new approach is proposed for detecting and tracking faint,small and moving multi-targets from a series images which are contaminated by high random noise and contain a complex background. The trajectories of moving objects are unknown,arbitrary but continuous. This paper presents a combined algorithm which consists of a moving contrast filter,AND operation process and the continuity filtering algorithm. This algorithm shows great effectiveness in detecting and tracking point targets from high-noisy and complex background image sequences. One of the properties of the algorithm is parallelism,and therefore it is suitable to the real time tracking tasks.