刚体碎块的断裂面匹配

发布时间：2018-07-16 11:48

【摘要】：目的刚体碎块匹配已经在考古、生物工程以及遥感数据处理等领域得到了较为广泛的应用,为了进一步提高碎块匹配的精度、速度和算法的抗噪性,提出一种先粗配再细配的刚体碎块匹配方法。方法首先采用基于显著性区域的碎块断裂面匹配方法实现碎块的粗匹配,然后通过加入高斯概率模型、角度约束和动态迭代系数的方式来改进迭代最近点(ICP)算法,并采用该算法来实现两个刚体碎块断裂面的细匹配,从而完成两个碎块的最终精确匹配。结果通过分别对公共碎块数据集和带有噪声的秦俑碎块数据模型的匹配实验结果表明,与ICP(iterative closest point)算法和概率迭代最近点(PICP)算法相比,提出的改进ICP算法在精度方面分别提高了约50%和15%,在速度方面分别提高了约65%和50%,是一种精度更高、速度更快、抗噪性更强的点集匹配算法。结论该方法不仅能够实现公共碎块数据集的完美匹配,而且对于秦俑这种特殊的刚体碎块也具有良好的匹配效果,会有更加广阔的应用领域和发展前景。
[Abstract]:Objective rigid fragment matching has been widely used in archaeology, bioengineering and remote sensing data processing. This paper presents a matching method of rigid body fragments with coarse matching and fine matching. Methods first of all, the coarse matching of fragments is realized by using the method of fault surface matching based on significant region, and then the iterative nearest point algorithm is improved by adding Gao Si probabilistic model, angle constraint and dynamic iteration coefficient. The algorithm is used to realize the fine matching of the fracture surface of the two rigid body fragments, so as to complete the final accurate matching of the two fragments. Results by matching the common data sets and the noisy terracotta Warriors data model, the experimental results show that, compared with the ICP (iterative closest point) algorithm and the probabilistic iterative nearest Point (PICP) algorithm, The improved ICP algorithm improves the accuracy by about 50% and 15%, and increases the speed by 65% and 50% respectively. It is a point set matching algorithm with higher accuracy, faster speed and stronger anti-noise. Conclusion this method can not only achieve the perfect matching of common fragment data sets, but also has a good matching effect for the special rigid body fragments such as the terracotta warriors, and will have a wider application field and development prospect.
【作者单位】：咸阳师范学院教育科学学院;西北大学信息科学与技术学院;北京师范大学信息科学与技术学院;
【基金】：国家自然科学基金项目(61373117) 陕西省教育厅科学研究项目(16jk2178)~~
【分类号】：TP391.41

【相似文献】