一种优化的火箭弹图像检测与跟踪算法

文章研究火箭弹图像的检测及跟踪。首先,提出了一种基于B样条函数对火箭弹图像进行边缘检测的算法,文中推导了相应的数学模型,并通过最大熵阈值法、图像细化以及计算中心点等,以提取火箭弹图像形心坐标,在此基础上提出一种基于最优组合模型对火箭弹运行参数进行跟踪,该模型的主要特点是避免了单一预测模型仅利用部分信息而带来的不足。文中对实际拍摄的火箭弹图像进行实验研究,得到的图像边缘清晰连续、中心点位置准确,较目前常用算法在火箭弹图像边缘检测方面有更大的优势,文中同时仿真证实了优化预测模型较单一预测方法降低了预测误差、提高了预测精度,有较好的应用前景。

An Optimal Algorithm for Detecting and Tracking Rocket

This paper studies an optical algorithm for detecting and tracking rocket.Based on B-spline wavelets,a novel algorithm of image edge detection is proposed,and the mathematical model is derived.By threshold estimating using maximum entropy,figure thinning and center calculating,the center point for rocket is achieved.For tracking rocket in real time,a forecasting model for rocket tracking using optimally combined model is presented.The main characteristic of using composite model instead of using only one method is that it can make use of most of the available factors so as to avoid the drawback of conventional forecasting model where only partial information is used.Experiments show noise is eliminated availably using this detecting algorithm and hereby the more precise position of target is gotten,this algorithm is superior to the current algorithms in terms of detection precision of the rocket image edges.And simulation results show that the forecasting model can reduce the prediction error and yield higher forecasting precision compared to using onlv one method.and is therefore an attractive scheme for practical systems.