基于多目标提取与航迹处理的目标告警算法

摘要

针对模板匹配目标告警算法计算量大、虚警率高的缺点,提出了基于多目标提取与航迹处理的目标告警算法.该算法采用“3×3中值滤波与图像增强”+“8向梯度检测目标提取”+“目标聚类”算法,提高目标提取的准确性；采用航迹处理算法降低目标告警的虚警率.试验表明:基于多目标提取与航迹处理的目标告警算法在减少计算量的同时,提高了目标告警的准确性,降低了目标告警的虚警率.%Because of template matching target alarm algorithm is of high false alarm rate and large calculation amount,a target alarm algorithm based on multi-target extraction and track processing was put forward. The algorithm used 3×3 median filtering and image enhancement + 8 direction gradient detection to extract target + target clustering algorithm to improve the accuracy of target extraction. Track processing algorithm reduced false alarm rate. The experiments showed that the target alarm algorithm based on multi-target extraction and track processing could reduc the amount of computation, improve the accuracy of the target alarm and reduce false a-larm rate.