改进Hausdorff距离和粒子群算法在激光制导中的应用

为提高激光成像制导精度,实现遮挡条件下的有效识别,提出一种基于改进Hausdorff距离和粒子群算法的激光图像匹配算法。首先提取基准图和实时图的边缘特征;而后针对原始Hausdorff距离易受噪声、孤立点及遮挡影响的不足,提出一种自适应部分均值Hausdorff距离,并将其作为相似性测度;最后改进粒子群算法以完成搜索匹配,一方面提出混沌惯性权值以提高其搜索能力,另一方面通过引入混沌局部搜索避免算法过早收敛。实验结果表明,该算法不仅具有较高的匹配成功率,而且实时性较好。

Application of improved Hausdorff distance and particle swarm optimization in laser imaging guidance

In order to improve the accuracy of laser imaging guidance and realize the effective recognition under occlusion conditions,a laser image matching algorithm based on improved Hausdorff distance and particle swarm optimization is proposed.Firstly,the edge features of the reference image and the real-time image were extracted.Then,because the original Hausdorff distance is susceptible to the noise,isolated point and occlusion,an adaptive partial mean Hausdorff distance is proposed as the similarity measure function.Finally,the particle swarm optimization is improved to complete the search matching.On the one hand,the chaotic inertia weight is proposed to improve the searching ability,on the other hand,the chaos local search is used to avoid the premature convergence.Experimental results show that this algorithm has a high matching success rates,and has good real-time.