Line Feature Extraction Research Articles

A Single LiDAR-Based Feature Fusion Indoor Localization Algorithm.

In past years, there has been significant progress in the field of indoor robot localization. To precisely recover the position, the robots usually relies on multiple on-board sensors. Nevertheless, this affects the overall system cost and increases computation. In this research work, we considered a light detection and ranging (LiDAR) device as the only sensor for detecting surroundings and propose an efficient indoor localization algorithm. To attenuate the computation effort and preserve localization robustness, a weighted parallel iterative closed point (WP-ICP) with interpolation is presented. As compared to the traditional ICP, the point cloud is first processed to extract corners and line features before applying point registration. Later, points labeled as corners are only matched with the corner candidates. Similarly, points labeled as lines are only matched with the lines candidates. Moreover, their ICP confidence levels are also fused in the algorithm, which make the pose estimation less sensitive to environment uncertainties. The proposed WP-ICP architecture reduces the probability of mismatch and thereby reduces the ICP iterations. Finally, based on given well-constructed indoor layouts, experiment comparisons are carried out under both clean and perturbed environments. It is shown that the proposed method is effective in significantly reducing computation effort and is simultaneously able to preserve localization precision.

Geometrical flow‐guided fast beamlet transform for crack detection

Beamlet transform has been widely used for extracting line features from images, which is an excellent multiscale geometric analysis method. However, it has a major drawback that it always performs too slowly due to very much redundant computation. In many application fields, the speed of the original beamlet transform is almost unbearable. To cure the problem, beamlet transform is improved by introducing geometrical flow, which utilises image semantic information in the process of generating beamlets. Besides, to further speed up the algorithm, interesting factor is presented to reduce recursively partitioned boxes. As a result, lots of computation time is saved. Experiments are conducted on various crack images and the results show that the proposed method runs significantly faster than the original beamlet transform. Cracks in an image are detected accurately. Moreover, the proposed method is robustly enough since the performance is hardly affected by crack shape and background texture.

基于区域聚类分割的点云特征线提取

基于区域聚类分割的点云特征线提取

APPLICATIONS OF 3D-EDGE DETECTION FOR ALS POINT CLOUD

Abstract. Edge detection has been one of the major issues in the field of remote sensing and photogrammetry. With the fast development of sensor technology of laser scanning system, dense point clouds have become increasingly common. Precious 3D-edges are able to be detected from these point clouds and a great deal of edge or feature line extraction methods have been proposed. Among these methods, an easy-to-use 3D-edge detection method, AGPN (Analyzing Geometric Properties of Neighborhoods), has been proposed. The AGPN method detects edges based on the analysis of geometric properties of a query point’s neighbourhood. The AGPN method detects two kinds of 3D-edges, including boundary elements and fold edges, and it has many applications. This paper presents three applications of AGPN, i.e., 3D line segment extraction, ground points filtering, and ground breakline extraction. Experiments show that the utilization of AGPN method gives a straightforward solution to these applications.

Random cutting plane approach for identifying volumetric features in a CAD mesh model

This paper presents a method to identify regions that make up features like holes, slots, pockets as well as interacting features in a three-dimensional mesh of a computer-aided design (CAD)/Engineering model. Feature recognition is an important area in the field of CAD/Engineering with applications in model retrieval, creating an analysis model by defeaturing of the designed model for finite element applications, etc. Most feature recognition methods use either a cluster-based decomposition or feature line extraction through solid angles or curvature values, followed by graph-based heuristics. Such approaches require a user parameter for clustering or a threshold value for angle/curvature, neither of which is an easily deterministic one. The proposed algorithm identifies the features using contours generated by random cutting planes, followed by graph traversals (and not using heuristics) and without using parameter/threshold values. The algorithm can identify blind holes, through holes, slots and pockets. The geometry of most of the extracted features has also been identified using Gauss map. Interacting features have also been extracted and separated, which normally pose difficulty for most algorithms. Extensive experiments on CAD models from various benchmarks show that the algorithm is robust. Comparison with different algorithms (of which code was available) shows that our approach performs admirably and in the case of interacting features, the algorithm performs better than the existing ones.

Terrain Feature Line Extraction by Improved Gradient-Based Profile Analysis

Terrain Feature Line Extraction by Improved Gradient-Based Profile Analysis

基于激光测距仪的旋翼飞行器位置估计算法

本文通过采用激光测距仪获取周围环境的信息，运用迭代最近点(Iterative Closest Point, ICP)算法计算出位置，并通过对扩展卡尔曼滤波(Extended Kalman Filter, EKF)算法的改进，减少EKF算法的计算量，使旋翼飞行器能更加快速的实现对位置的估计。该算法通过在预测更新的过程中引进ICP算法计算的结果构建过程模型，并在测量更新过程中通过对环境中线段特征的提取，构建观测模型，结合飞控能够提供较为精确的角度信息的特点，简化观测模型的表达式，最终得到准确的位置信息。 In this paper, the environment information is captured and the position of rotorcraft is calculated by using laser scanner and using iterative closest point (ICP) algorithm. It could estimate the posi-tion more rapid and accurate by improving the extended Kalman filter (EKF) algorithm which re-duces the amount of computation. The trajectory model of rotorcraft will be constructed by com-bining and fusing the ICP and improved EKF algorithms to update the prediction. In the process of updating, the observed environment model will be sketched through extracting line features for surrounding of the rotorcraft. Then the model expression is simplified according to the accurate attitude angles of rotorcraft which come from Inertial Measurement Unit (IMU) sensor. Finally, location information is estimated accurately for the rotorcraft via simulation and experiment.

Shortwave Infrared Imaging Spectroscopy for Analysis of Ancient Paintings.

Spectral analysis is one of the main non-destructive techniques used to examine cultural relics. Hyperspectral imaging technology, especially on the shortwave infrared (SWIR) band, can clearly extract information from paintings, such as color, pigment composition, damage characteristics, and painting techniques. All of these characteristics have significant scientific and practical value in the study of ancient paintings and other relics and in their protection and restoration. In this study, an ancient painting, numbered Gu-6541, which had been found in the Forbidden City, served as a sample. A ground-based SWIR imaging spectrometer was used to produce hyperspectral images with high spatial and spectral resolution. Results indicated that SWIR imaging spectral data greatly facilitates the extraction of line features used in drafting, even using a single band image. It can be used to identify and classify mineral pigments used in paintings. These images can detect alterations and traces of daub used in painting corrections and, combined with hyperspectral data analysis methods such as band combination or principal component analysis, such information can be extracted to highlight outcomes of interest. In brief, the SWIR imaging spectral technique was found to have a highly favorable effect on the extraction of line features from drawings and on the identification of colors, classification of paintings, and extraction of hidden information.

EXTRACTION OF BUILDING BOUNDARY LINES FROM AIRBORNE LIDAR POINT CLOUDS

Abstract. Building boundary lines are important spatial features that characterize the topographic maps and three-dimensional (3D) city models. Airborne LiDAR Point clouds provide adequate 3D spatial information for building boundary mapping. However, information of boundary features contained in point clouds is implicit. This study focuses on developing an automatic algorithm of building boundary line extraction from airborne LiDAR data. In an airborne LiDAR dataset, top surfaces of buildings, such as roofs, tend to have densely distributed points, but vertical surfaces, such as walls, usually have sparsely distributed points or even no points. The intersection lines of roof and wall planes are, therefore, not clearly defined in point clouds. This paper proposes a novel method to extract those boundary lines of building edges. The extracted line features can be used as fundamental data to generate topographic maps of 3D city model for an urban area. The proposed method includes two major process steps. The first step is to extract building boundary points from point clouds. Then the second step is followed to form building boundary line features based on the extracted boundary points. In this step, a line fitting algorithm is developed to improve the edge extraction from LiDAR data. Eight test objects, including 4 simple low buildings and 4 complicated tall buildings, were selected from the buildings in NCKU campus. The test results demonstrate the feasibility of the proposed method in extracting complicate building boundary lines. Some results which are not as good as expected suggest the need of further improvement of the method.

ORIENTATION MODELING FOR AMATEUR CAMERAS BY MATCHING IMAGE LINE FEATURES AND BUILDING VECTOR DATA

Abstract. With the popularity of geospatial applications, database updating is getting important due to the environmental changes over time. Imagery provides a lower cost and efficient way to update the database. Three dimensional objects can be measured by space intersection using conjugate image points and orientation parameters of cameras. However, precise orientation parameters of light amateur cameras are not always available due to their costliness and heaviness of precision GPS and IMU. To automatize data updating, the correspondence of object vector data and image may be built to improve the accuracy of direct georeferencing. This study contains four major parts, (1) back-projection of object vector data, (2) extraction of image feature lines, (3) object-image feature line matching, and (4) line-based orientation modeling. In order to construct the correspondence of features between an image and a building model, the building vector features were back-projected onto the image using the initial camera orientation from GPS and IMU. Image line features were extracted from the imagery. Afterwards, the matching procedure was done by assessing the similarity between the extracted image features and the back-projected ones. Then, the fourth part utilized line features in orientation modeling. The line-based orientation modeling was performed by the integration of line parametric equations into collinearity condition equations. The experiment data included images with 0.06 m resolution acquired by Canon EOS Mark 5D II camera on a Microdrones MD4-1000 UAV. Experimental results indicate that 2.1 pixel accuracy may be reached, which is equivalent to 0.12 m in the object space.

ORIENTATION MODELING FOR AMATEUR CAMERAS BY MATCHING IMAGE LINE FEATURES AND BUILDING VECTOR DATA

With the popularity of geospatial applications, database updating is getting important due to the environmental changes over time. Imagery provides a lower cost and efficient way to update the database. Three dimensional objects can be measured by space intersection using conjugate image points and orientation parameters of cameras. However, precise orientation parameters of light amateur cameras are not always available due to their costliness and heaviness of precision GPS and IMU. To automatize data updating, the correspondence of object vector data and image may be built to improve the accuracy of direct georeferencing. This study contains four major parts, (1) back-projection of object vector data, (2) extraction of image feature lines, (3) object-image feature line matching, and (4) line-based orientation modeling. In order to construct the correspondence of features between an image and a building model, the building vector features were back-projected onto the image using the initial camera orientation from GPS and IMU. Image line features were extracted from the imagery. Afterwards, the matching procedure was done by assessing the similarity between the extracted image features and the back-projected ones. Then, the fourth part utilized line features in orientation modeling. The line-based orientation modeling was performed by the integration of line parametric equations into collinearity condition equations. The experiment data included images with 0.06 m resolution acquired by Canon EOS Mark 5D II camera on a Microdrones MD4-1000 UAV. Experimental results indicate that 2.1 pixel accuracy may be reached, which is equivalent to 0.12 m in the object space.

Extracting Line Features in SAR Images Through Image Edge Fields

Conventional line detection methods are mainly based on the binary edge map. This letter proposes a new line detection method that directly extracts line features from the image edge fields of the synthetic aperture radar (SAR) images. In the proposed method, the strength and direction of each field point are first obtained using a ratio-based edge filter. Then, the accumulation weight of the field point is jointly computed using its strength and direction. The direction of a field point on the line is essentially the orientation of the line. Furthermore, a field point on a strong line should be distinguished from a field point on a weak line. Thus, the accumulation weights of different field points are not equal. By summing up the accumulation weights, the straight lines in the SAR image space are directly converted into several local peaks in the parameter space. A sort-window peak detection method is proposed to suppress the spurious secondary peaks in the parameter space. The experimental results show that the proposed line detection method is robust to noise and has a good antiocclusion ability. The proposed method performs well in terms of true positive detection rate and detection accuracy for both synthetic and real-world images.

Geometric Calibration of Micro-Lens-Based Light Field Cameras Using Line Features.

We present a novel method for the geometric calibration of micro-lens-based light field cameras. Accurate geometric calibration is the basis of various applications. Instead of using sub-aperture images, we directly utilize raw images for calibration. We select appropriate regions in raw images and extract line features from micro-lens images in those regions. For the entire process, we formulate a new projection model of a micro-lens-based light field camera, which contains a smaller number of parameters than previous models. The model is transformed into a linear form using line features. We compute the initial solution of both the intrinsic and the extrinsic parameters by a linear computation and refine them via non-linear optimization. Experimental results demonstrate the accuracy of the correspondences between rays and pixels in raw images, as estimated by the proposed method.

Extracting feature lines from point clouds based on smooth shrink and iterative thinning

An index of measuring the variation on a surface called the smooth shrink index (SSI) which presents robustness to noise and non-uniform sampling is developed in this work. Afterwards, a new algorithm used for extracting the feature lines was proposed. Firstly, the points with an absolute value of SSI greater than a given threshold are selected as potential feature points. Then, the SSI is applied as the growth condition to conduct region segmentation of the potential feature points. Finally, a bilateral filter algorithm is employed to obtain the final feature points by thinning the potential feature points iteratively. While thinning the potential feature points, the tendency of the feature lines is acquired using principle component analysis (PCA) to restrict the drift direction of the potential feature points, so as to prevent the shrink in the endpoints of the feature lines and breaking of the feature lines induced by non-uniform sampling.

A statistical approach for extraction of feature lines from point clouds

This paper firstly introduces the use of a statistical model based on the Poisson distribution as a tool for extracting feature points from point clouds. When the features on a model are non-uniform or the surfaces are not completely smooth, the thresholds for different local features must be set differently. Such case is ill-suited to all previous global fixed thresholds dependent methods. The Poisson distribution based method is preferable because it computes different thresholds for different local features adaptively depending on the natural properties of the surfaces. Secondly, the region information analysis is employed to cluster the border/edge points, identify the corner points and acquire the link information for feature lines. Finally, the complete feature lines are reconstructed based on the geometry representations of the border point clusters that are created by adapting L1-median locally, instead of using distance/path parameters. The proposed method does not need any prior surface reconstruction and is not largely affected by noisy points, neighborhood scales or sampling quality. We demonstrate the benefits of our method with the favorable results for real-world point clouds with varying types of features and the application of the generated feature lines to the computer aided line-drawings generation of Terracotta.

TRANSFORMASI HOUGH LINEAR UNTUK ANALISIS DAN PENGENALAN BATIK MOTIF PARANG

Batik is a craft that has high artistic value. Batik also has become part of Indonesian culture (especially Java) since long. There are so many designs old patterns of Batik. One of traditional batik motif is Parang. Batik Parang have decorative striped pattern and lined tilted. This research analizes characteristic of batik Parang and develops a software to automatically identify motifs of batik Parang image using line feature extraction. The pattern can be identified through a group of points that form the edge of a line and then detected as a line using Hough transform. The research material was 50 image of the sample data and 30 image of research datarespectivelyParang and non Parang. Both the accuracy of Batik Parang and non Parang recognition are 90%.

Automatic registration of large-scale urban scene point clouds based on semantic feature points

Point clouds collected by terrestrial laser scanning (TLS) from large-scale urban scenes contain a wide variety of objects (buildings, cars, pole-like objects, and others) with symmetric and incomplete structures, and relatively low-textured surfaces, all of which pose great challenges for automatic registration between scans. To address the challenges, this paper proposes a registration method to provide marker-free and multi-view registration based on the semantic feature points extracted. First, the method detects the semantic feature points within a detection scheme, which includes point cloud segmentation, vertical feature lines extraction and semantic information calculation and finally takes the intersections of these lines with the ground as the semantic feature points. Second, the proposed method matches the semantic feature points using geometrical constraints (3-point scheme) as well as semantic information (category and direction), resulting in exhaustive pairwise registration between scans. Finally, the proposed method implements multi-view registration by constructing a minimum spanning tree of the fully connected graph derived from exhaustive pairwise registration. Experiments have demonstrated that the proposed method performs well in various urban environments and indoor scenes with the accuracy at the centimeter level and improves the efficiency, robustness, and accuracy of registration in comparison with the feature plane-based methods.

Clustering Based Line Detection in Noisy Datasets

For autonomous navigation of a mobile robot in indoor environments, accurate mapping of indoor structures is a crucial requirement. When robots are operating in a noisy cluttered environments the noise generated from the external sensors affect extracting features from sensor data. In this work, we present a novel method of extracting line segments from LIDAR data in very noisy datasets to accomplish accurate mapping. As compared to traditional methods of extracting line features from sensor data, our method can generate very accurate line features even in a very cluttered areas. We present a novel hybrid algorithm that uses Hough transform and clustering techniques that is fast and efficient.

3D reconstruction of building facade with fused data of terrestrial LiDAR data and optical image

We describe a novel method of data fusion of terrestrial LiDAR and images for delineating and deriving features and 3D reconstruction of regular building facades. We discuss the registration of LiDAR and image based on the transferring parameters of coordinate systems. We conduct the initial features extraction of building from images based on the gradient direction, and the further processing of the features constrained by the integrity rules of building construction. And consequently we implement the converting of 2D image feature lines’ to 3D point cloud space based on the former registration, and then the 3D point cloud data around the 3D feature lines are searched. Finally we obtain the more accurate 3D feature lines after a series of operations of “buffer area – plane fitting – feature line extraction” on the point cloud data around the feature lines, which is for better presentation of building model. At the end of the paper, the experiment of the data fusion of image and LiDAR was conducted. We argue that the method represent a fundamental and useful approach to improve and explore the accuracy and efficiency of 3D reconstruction of building façade based on terrestrial LiDAR data and optical image.

Development of kinematic 3D laser scanning system for indoor mapping and as-built BIM using constrained SLAM.

The growing interest and use of indoor mapping is driving a demand for improved data-acquisition facility, efficiency and productivity in the era of the Building Information Model (BIM). The conventional static laser scanning method suffers from some limitations on its operability in complex indoor environments, due to the presence of occlusions. Full scanning of indoor spaces without loss of information requires that surveyors change the scanner position many times, which incurs extra work for registration of each scanned point cloud. Alternatively, a kinematic 3D laser scanning system, proposed herein, uses line-feature-based Simultaneous Localization and Mapping (SLAM) technique for continuous mapping. Moreover, to reduce the uncertainty of line-feature extraction, we incorporated constrained adjustment based on an assumption made with respect to typical indoor environments: that the main structures are formed of parallel or orthogonal line features. The superiority of the proposed constrained adjustment is its reduction for uncertainties of the adjusted lines, leading to successful data association process. In the present study, kinematic scanning with and without constrained adjustment were comparatively evaluated in two test sites, and the results confirmed the effectiveness of the proposed system. The accuracy of the 3D mapping result was additionally evaluated by comparison with the reference points acquired by a total station: the Euclidean average distance error was 0.034 m for the seminar room and 0.043 m for the corridor, which satisfied the error tolerance for point cloud acquisition (0.051 m) according to the guidelines of the General Services Administration for BIM accuracy.