共查询到20条相似文献,搜索用时 140 毫秒
1.
针对传统随机采样一致性(random sample consensus,RANSAC)算法对复杂屋顶的分割缺陷,提出一种复杂屋顶平面的RANSAC优化分割方法。首先,在点云法向量估计的基础上,利用点法式方程优化种子点选取过程,提高初始平面选择的有效性;然后,采用距离与法向加权法抑制虚假平面生成,并利用加权函数进行平面内点的迭代修正,提高面片分割的准确性;最后,利用面片竞争方法优化分割结果,实现屋顶点云的分割处理。多组点云分割实验结果表明,该方法能有效抑制虚假平面,对复杂建筑物的屋顶平面分割结果的精确率、召回率和整体精度分别达到96.8%、98.2%和95.1%。相比传统方法,该方法在点云分割结果正确率及耗时方面均有明显优势。 相似文献
2.
3.
建筑物屋顶面大小差异较大、形状复杂、数量不确定等特点,以及机载LiDAR点云密度不均、分布不规则、缺乏语义信息等特性,对屋顶面的准确分割造成了很大干扰,因此现有分割方法的精度和适用性仍有待提高.针对上述问题,本文提出一种结合区域增长与RANSAC的机载LiDAR点云屋顶面分割方法.首先,引入稳健的法向量估计算法计算点云法向量,利用提出的迭代区域增长策略和RANSAC提取多个可靠屋顶面片;然后,基于可靠屋顶面片参数和RANSAC计算内点的思想,迭代合并可靠屋顶面片,并精化屋顶面参数;最后,计算未能通过前面步骤分割的点到各屋顶面的垂直距离,将其标记为距离最小且小于阈值的屋顶面,并通过局部范围内投票的方式精化屋顶面分割结果.利用多个具有代表性的建筑物点云和一组区域建筑物点云进行试验,结果表明,所提出的方法可有效地分割不同复杂程度的建筑物屋顶面,并能较好地分割面积较小的屋顶面,以屋顶面和单点为评价单元的平均分割正确率为95.56% 和97.93%,分割的结果可为建筑物三维模型重建、点云精简等应用提供可靠的信息. 相似文献
4.
5.
6.
建筑物屋顶面点云分割结果的好坏对建筑物三维模型重建起着重要的作用。针对传统RANSAC算法建筑物屋顶面点云的分割问题,提出了一种基于局部约束的建筑物点云平面分割方法。利用点云局部曲面法向约束构建法向准则,利用半径约束的点云空间聚类的方法对共面屋顶面点云进行分解,从而抑制"伪屋顶面"的产生;利用局部抽样策略降低算法的迭代次数,减少运算量。实验表明该方法能够获得稳定可靠的建筑物屋顶点云分割结果,将有利于后续的建筑物三维模型重建。 相似文献
7.
8.
9.
针对车载激光扫描获取的道路点云数据分类问题多的难点,本文提出了一种基于最大类间方差(Otsu)算法与改进区域生长算法的道路面提取方法。原始点云中非地面点滤除依靠Otsu算法自适应计算出分割阈值;随后分别计算点云的法向量与曲率;最终将法向量相似度作为约束条件,使用改进区域生长算法进行道路面精确提取。通过两段典型的城市道路点云数据为例,试验结果表明,本文方法提取道路面结果的准确度(CR)、完整度(CP)以及提取质量(Q)均大于94%,充分证明了该方法的有效性。 相似文献
10.
提出一种基于图割算法的建筑物LiDAR点云与正射影像融合提取方法。首先,利用LiDAR点云计算3个几何特征:平整度、法向量分布和高程纹理一致性。同时利用航空正射影像计算颜色特征——归一化植被指数(NDVI)。然后将两类特征联合构建能量函数数据项,综合数字表面模型(DSM)和NDVI构建平滑项,采用图割算法优化得到初始的建筑物区域。最后利用初始建筑物边缘一定范围内的正射影像颜色信息,采用前后景分割的思想进一步优化建筑物边缘。应用ISPRS Vaihingen测试数据进行试验,结果表明本文方法具有较高的建筑物提取精度。 相似文献
11.
The extraction of object features from massive unstructured point clouds with different local densities, especially in the presence of random noisy points, is not a trivial task even if that feature is a planar surface. Segmentation is the most important step in the feature extraction process. In practice, most segmentation approaches use geometrical information to segment the 3D point cloud. The features generally include the position of each point (X, Y and Z), locally estimated surface normals and residuals of best fitting surfaces; however, these features could be affected by noisy points and in consequence directly affect the segmentation results. Therefore, massive unstructured and noisy point clouds also lead to bad segmentation (over-segmentation, under-segmentation or no segmentation). While the RANSAC (random sample consensus) algorithm is effective in the presence of noise and outliers, it has two significant disadvantages, namely, its efficiency and the fact that the plane detected by RANSAC may not necessarily belong to the same object surface; that is, spurious surfaces may appear, especially in the case of parallel-gradual planar surfaces such as stairs. The innovative idea proposed in this paper is a modification for the RANSAC algorithm called Seq-NV-RANSAC. This algorithm checks the normal vector (NV) between the existing point clouds and the hypothesised RANSAC plane, which is created by three random points, under an intuitive threshold value. After extracting the first plane, this process is repeated sequentially (Seq) and automatically, until no planar surfaces can be extracted from the remaining points under the existing threshold value. This prevents the extraction of spurious surfaces, brings an improvement in quality to the computed attributes and increases the degree of automation of surface extraction. Thus the best fit is achieved for the real existing surfaces. 相似文献
12.
本文利用地基激光雷达实现天然林区近地面点云数据的精细分类和倒木提取。对大兴安岭天然林区的3个倒木样地进行了近地面1.3 m以内点云精细分类和倒木信息提取。为避免点云密度差异和遮挡的形态特征,点云分类时基于自适应临近搜索法计算团块协方差特征值构造3D和2D特征。使用k临近递增的团块协方差特征值得到的线性特征、面状特征和发散状特征构造最大熵函数,用最大熵函数取得最大值时的临近点云计算特征参数,根据递归特征排除法(RFE)筛选重要变量进行随机森林分类。利用自适应kNN特征得到3块研究样地(A、B、C)的分类总体精度分别为93.17%、94.52%、95.16%;固定k临近搜索时,总体精度分别为92.65%、89.09%、92.99%,表明自适应kNN搜索方法使分类精度有一定提高。提取倒木点云去噪处理后进行随机抽样一致圆柱拟合,根据轴线方向进行圆柱的筛选与合并,实现倒木的识别,样地倒木识别率为100%。 相似文献
13.
14.
15.
Nusret Demir 《Journal of the Indian Society of Remote Sensing》2018,46(8):1265-1272
The purpose of this study is to derive vectoral 3D roof planes from the LIDAR point cloud of the detected buildings. For segmentation of the LIDAR point cloud, the RANSAC algorithm has been used. Because the RANSAC algorithm is sensitive to the used parameters, and results in over- or under-segmentation of the clusters, a refinement method has been proposed. The detection of roof planes has been improved with use of the refinement method. Therefore, similar plane surfaces have been combined, followed by the region-growing algorithm, to split the under-segmented plane surfaces. The digitization of the roof boundaries is performed using the alpha-shapes algorithm, followed by line fitting to generalize the roof edges. The quality assessment has been done using the reference vector dataset with comparison using four different criteria. 相似文献
16.
This paper addresses the problem of 3D surface scan refinement, which is desirable due to noise, outliers, and missing measurements being present in the 3D surfaces obtained with a laser scanner. We present a novel algorithm for the fusion of absolute laser scanner depth profiles and photometrically estimated surface normal data, which yields a noise-reduced and highly detailed depth profile with large scale shape robustness. In contrast to other approaches published in the literature, the presented algorithm (1) regards non-Lambertian surfaces, (2) simultaneously computes surface reflectance (i.e. BRDF) parameters required for 3D reconstruction, (3) models pixelwise incident light and viewing directions, and (4) accounts for interreflections. The algorithm as such relies on the minimization of a three-component error term, which penalizes intensity deviations, integrability deviations, and deviations from the known large-scale surface shape. The solution of the error minimization is obtained iteratively based on a calculus of variations. BRDF parameters are estimated by initially reducing and then iteratively refining the optical resolution, which provides the required robust data basis. The 3D reconstruction of concave surface regions affected by interreflections is improved by compensating global illumination in the image data. The algorithm is evaluated based on eight objects with varying albedos and reflectance behaviors (diffuse, specular, metallic). The qualitative evaluation shows a removal of outliers and a strong reduction of noise, while the large scale shape is preserved. Fine surface details Which are previously not contained in the surface scans, are incorporated through using image data. The algorithm is evaluated with respect to its absolute accuracy using two caliper objects of known shape, and based on synthetically generated data. The beneficial effect of interreflection compensation on the reconstruction accuracy is evaluated quantitatively in a Photometric Stereo framework. 相似文献
17.
18.
19.
针对目前传统图像匹配算法在复杂环境下存在误匹配点对过多、稳健性较差等问题,本文提出一种基于改进FAST的特征点提取,结合对立颜色特征的图像匹配算法。首先,利用改进FAST算法提取的角点作为特征点,结合改进的Opponent SIFT算法对特征点进行描述;然后,使用基于字符定位算法对提取的特征点对进行粗匹配,降低整体匹配过程中特征点对误匹配的风险。最后,为了规避因RANSAC算法易陷入局部最优解而导致正确点对被误剔除的问题,运用向量场一致性替代RANSAC进行提纯,降低误匹配率。通过对比试验表明,改进算法匹配准确率均大于91%,且对差异变化具有较好的稳健性、适应性。 相似文献
20.
地面三维激光扫描点云密集,实体信息隐含,特征提取困难。根据原始点云数据的特点,提出对点云进行实体化处理,种子点生成技术可用于实体提取。完成点云实体化后,通过随机采样一致算法结合稳健估计方法,可更好地进行地物特征提取。由于现实客观世界的复杂性,采用一些先验性知识,可以减少人工处理过程,加快点云分类和实体建模速度。 相似文献