Skeletonization of Plant Point Cloud Data Using Stochastic Optimization Framework

In this paper, a method of reconstructing 3D (three dimensional) models from the original scanned point cloud using priori templates is proposed. Different from previous reconstruction methods that triangulate and fit the original scanning point cloud directly, we construct a priori template based on the CAD (computer aided design) model and guide the reconstruction of the original scanning point cloud with the priori template. Given a CAD model, the basic geometric elements are used as the basic units to extract the set elements of 3D shapes. Then the geometric elements are meshed, and the normal vectors at the mesh nodes are extracted. The corresponding point cloud data of each basic element are extracted from the original point cloud. The point cloud data near the normal of the guide point are searched, and the Gaussian weighted average value of the searched point represents the actual geometric parameters of the part at the guide point. Finally, the geometric elements of the basic unit are reconstructed locally by Non-Uniform Rational B-Splines surface fitting, and the complete reconstruction model is obtained by integrating the local reconstruction. Experiments show that our method can solve the problems of high quality reconstruction, sharp feature preservation, and detail recovery in surface reconstruction.

A hybrid method for skeleton extraction on Kinect sensor data: Combination of L1-Median and Laplacian shrinking algorithms

As one of the most important oil crops, the high-accuracy rapeseed digital model contributes to agronomic horticulture, crop-breeding and plant genetics research. How to build a three-dimensional (3D) plant model, especially a high-quality leaf model, has great significance to the research on the rapeseed plant growth status. In this paper, we present a three-dimensional oilseed rape plant reconstruction method based on point cloud data. Firstly, we provide a novel approach to reconstruct 3D rapeseed shape model by utilizing the acquired 3D point clouds, and the original point cloud acquired via 3D laser scanning should be preprocessed, the main idea of processing includes de-noising and simplifying the original point clouds. Then the meshing method and mesh optimization process have been carried out to attain a high-quality 3D rapeseed plant mesh model. Thus the rapeseed plants can be non-destructively monitored by measuring the reconstructed rapeseed plant models, and we can obtain the needed parameters such as the leaf area. Experiments are implemented on real datasets, and the results indicate that the accuracy of the leaf area values are highly closed to the real values, which are measured by conventional digital processing methodology.

Numerical modelling of airflow in underground mines is gaining importance in modern ventilation system design and safety assessment. Computational Fluid Dynamics (CFD) simulations enable detailed analyses of air movement, contaminant dispersion, and heat transfer, yet their reliability depends strongly on the accuracy of the geometric representation of excavations. Raw point cloud data obtained from laser scanning of underground workings are typically irregular, noisy, and contain discontinuities that must be processed before being used for CFD meshing. This study presents a methodology for automatic segmentation and regularization of large-scale point cloud data of underground excavation systems. The proposed approach is based on skeleton extraction and trajectory analysis, which enable the separation of excavation networks into individual tunnel segments and crossings. The workflow includes outlier removal, alpha-shape generation, voxelization, medial-axis skeletonization, and topology-based segmentation using neighbor relationships within the voxel grid. A proximity-based correction step is introduced to handle doubled crossings produced by the skeletonization process. The segmented sections are subsequently regularized through radial analysis and surface reconstruction to produce uniform and watertight models suitable for mesh generation in CFD software (Ansys 2024 R1). The methodology was tested on both synthetic datasets and real-world laser scans acquired in underground mine conditions. The results demonstrate that the proposed segmentation approach effectively isolates single-line drifts and crossings, ensuring continuous and smooth geometry while preserving the overall excavation topology. The developed method provides a robust preprocessing framework that bridges the gap between point cloud acquisition and numerical modelling, enabling automated transformation of raw data into CFD-ready geometric models for ventilation and safety analysis of complex underground excavation systems.

Classification and recognition of objects for images is an important issue in many scientific fields such as computer vision, biometrics and medical image analysis. An important feature of many objects is shape, so shape analysis has become an important part of classification. One method of shape analysis is to estimate boundaries and analyze the shape of these boundaries to determining the characteristics of the original object. However, many literature studies on point cloud shape analysis are based on existing shapes. This paper mainly refers to Ripley’s K-function in spatial point analysis, through this judgment on spatial distribution of point cloud data to determine the existence of shape in point cloud data, through the spatial distribution of 2D point cloud data and 3D point cloud data. Judging by the randomness of the experiment, K-function has a considerable effect on judging existence of point cloud data shape through relevant experimental verification analysis.

Learning on (3D) point clouds is vital for a broad range of emerging applications such as autonomous driving, robot perception, augmented reality, gaming, and security. Such needs have increased recently due to the prevalence of 3D sensors such as LiDAR, 3D camera, and RGB-D. Point clouds consist of thousands to millions of points; They contain rich information and are complementary to the traditional 2D cameras that we have been working on for years in the multimedia (or vision) community. 3D learning algorithms on point cloud data are new, and exciting, for numerous core problems such as 3D classification, detection, semantic segmentation, and face recognition. Covers the requirements of point cloud data, the background of capturing the data, 3D representations, emerging applications, core problems, state-of-the art learning algorithms, and future research opportunities.

Point cloud is the popular researches in the industry, and the research on partial point cloud is favored by many researchers. In the acquisition process of original 3D point cloud data, there are partial or sparse problems due to occlusion, lighting and other reasons, which will lead to deviation in downstream tasks. We propose a network MDPCN using multiple decoders, which can make the partial point cloud complete. The network uses deep learning to get multiple features of the partial point cloud, and uses multiple identical decoders to decode. Each decoder uses the self-attention module and the Folding-Net to generate smooth point clouds from the features, and finally integrates the point clouds generated by each decoder to obtain dense point cloud. Compared with other mainstream point cloud completion network methods and ablation experiments, it can be proved that this network model can generate more accurate point cloud and effectively complete the 3D reconstruction task.

One effective iterative method for estimating parameters in latent variable probabilistic models is the Expectation-Maximization (EM) algorithm, which finds extensive use in classification issues. A versatile probabilistic framework for modeling data from multiple Gaussian distributions, Gaussian Mixture Models (GMMs) employ EM for parameter estimation, which is explored in this article. Problems with class membership estimation (including means, covariances, and mixing coefficients) are the primary focus of this study. Starting with initial parameter estimations, estimating posterior probabilities in the E-step, and repeatedly updating model parameters in the M-step until convergence, the technique entails applying the EM algorithm on both synthetic and real-world datasets. In order to measure performance, we compare it to K-means and other baseline classifiers using measures like classification accuracy, Adjusted Rand Index (ARI), and log-likelihood progression. When data is distributed according to a multimodal Gaussian distribution, the findings show that EM-GMM provides better parameter estimation and better classification accuracy. Initialization procedures have a substantial effect on performance, as convergence study further shows. In addition to reiterating the EM algorithm's usefulness for probabilistic classification, this paper also describes its advantages and disadvantages and offers suggestions for improving its performance in settings with high-dimensional data and noise.

3D point cloud data captured from Oil and Gas (O&G) facilities are becoming increasingly abundant. This data supports the interactive visualization of complex plants, which is central to numerous industrial advancements. As industries seek to innovate and optimize their processes, the versatile use of 3D point cloud data is proving invaluable beyond traditional O&G applications. The precision and comprehensiveness of 3D point cloud data facilitate efficient design, construction, maintenance and operational monitoring across various industrial sectors. Digital Twin (DT) solutions enable this by maintaining 3D representations, such as 3D CAD models, that accurately reflect the current state of facilities. As installations change, the process of 3D scanning is often repeated to capture the latest updates. Engineers and designers then use modelling software to visualize the resulting point cloud and update the facility's 3D representation, incorporating any new modifications. However, this involves interpreting the visual point cloud data and structuring it into meaningful objects (such as valves, tanks, etc.), a process that remains manual, requiring labor-intensive tasks that are prone to errors and incur significant costs due to the need for specialized labor. We propose an Artificial Intelligence approach, that uses a deep learning-based algorithm to automatically segment 3D objects within a point cloud. Our workflow recognizes instances of usual components of production systems, such as control valves, pipes, and transmitters. A state-of-the-art 3D instance segmentation approach is adapted to the O&G domain. Given the lack of publicly available instance segmentation datasets in this domain, we manually annotate 54 scenes from an O&G offshore plant, each of them around 20 m2 in footprint. To further enhance model training, we also incorporated synthetic point clouds. Practical results on real point clouds, acquired from operational facilities and pre-processed as required by commercial DT solutions, indicated a performance of approximately 70% success (f1-score) in segmenting components of O&G production systems, reaching up to 85% for components such as control valves. Accordingly, a 52.4% reduction in costs and efforts associated with manually interpreting 3D point clouds is estimated. Point cloud instance segmentation has attracted much attention in the computer vision research community, and many approaches have been proposed. However, results in the literature are heavily biased toward a few benchmarking datasets, which were acquired in indoor environments like offices and rooms. Few studies have addressed the O&G domain, creating a gap in the literature for achieving more competitive results. This lack of comprehensive research limits the potential for innovation and optimization in the industry. This paper contributes by (i) assessing the performance of state-of-the-art algorithms in the domain of O&G facilities, (ii) proposing modifications to better generalize to components of O&G production systems, (iii) objectively measuring the impact of this algorithm in reducing manual effort for engineers, and (iv) advancing towards a new DT functionality that derives an asset 3D representation, that is faithful to physical reality, through the automatic interpretation of point cloud scans.

Abstract. Filtering is a key step for most applications of airborne LiDAR point clouds. Although lots of filtering algorithms have been put forward in recent years, most of them suffer from parameters setting or thresholds adjusting, which will be time-consuming and reduce the degree of automation of the algorithm. To overcome this problem, this paper proposed a threshold-free filtering algorithm based on expectation-maximization. The proposed algorithm is developed based on an assumption that point clouds are seen as a mixture of Gaussian models. The separation of ground points and non-ground points from point clouds can be replaced as a separation of a mixed Gaussian model. Expectation-maximization (EM) is applied for realizing the separation. EM is used to calculate maximum likelihood estimates of the mixture parameters. Using the estimated parameters, the likelihoods of each point belonging to ground or object can be computed. After several iterations, point clouds can be labelled as the component with a larger likelihood. Furthermore, intensity information was also utilized to optimize the filtering results acquired using the EM method. The proposed algorithm was tested using two different datasets used in practice. Experimental results showed that the proposed method can filter non-ground points effectively. To quantitatively evaluate the proposed method, this paper adopted the dataset provided by the ISPRS for the test. The proposed algorithm can obtain a 4.48 % total error which is much lower than most of the eight classical filtering algorithms reported by the ISPRS.

With the rapid advancement in laser technology, computer vision, and embedded computing, the application of laser scanning to the digitization of three dimensional physical realities has become increasingly widespread. In this paper, we focus on research results embodied in a 3D human body color digitization system developed at Tianjin University, and in collaboration with the Hong Kong University of Science and Technology. In digital human modeling, the first step involves the acquisition of the 3D human body data. We have over the years developed laser scanning technological know-how from first principles to support our research activities on building the first 3D digital human database for ethnic Chinese. The disadvantage of the conventional laser scanning is that surface color information is not contained in the point cloud data. By adding color imaging sensors to the developed multi-axis laser scanner, both the 3D human body coordinate data and the body surface color mapping are acquired. Our latest development is focused on skeleton extraction which is the key step towards human body animation, and applications to dynamic anthropometry. For dynamic anthropometric measurements, we first use an animation algorithm to adjust the 3D digital human to the required standard posture for measurement, and then fix the feature points and feature planes based on human body geometric characteristics. Utilizing the feature points, feature planes, and the extracted human body skeleton, we have measured 40 key sizes for the stand posture, and the squat posture. These experimental results will be given, and the factors that affect the measurement precision are analyzed through qualitative and quantitative analyses.

With the use of terrestrial laser scanning (TLS) in forest stands, surveys are now equipped to obtain dense point cloud data. However, the data range, i.e., the number of points, often reaches the billions or even higher, exceeding random access memory (RAM) limits on common computers. Moreover, the processing time often also extends beyond acceptable processing lengths. Thus, in this paper, we present a new method of efficiently extracting stem traits from huge point cloud data obtained by TLS, without subdividing or downsampling the point clouds. In this method, each point cloud is converted into a wireframe model by connecting neighboring points on the same continuous surface, and three-dimensional points on stems are resampled as cross-sectional points of the wireframe model in an out-of-core manner. Since the data size of the section points is much smaller than the original point clouds, stem traits can be calculated from the section points on a common computer. With the study method, 1381 tree stems were calculated from 3.6 billion points in ~20 min on a common computer. To evaluate the accuracy of this method, eight targeted trees were cut down and sliced at 1-m intervals; actual stem traits were then compared to those calculated from point clouds. The experimental results showed that the efficiency and accuracy of the proposed method are sufficient for practical use in various fields, including forest management and forest research.

Continuation of point clouds via persistence diagrams

We introduce a distance field guided \(L_1\)-median method to extract topologically clean 1D curve skeleton from the point cloud model. We first voxelize the input point cloud, and compute the distance field for the point cloud. Then with the distance field, we extract the initial skeleton of the model using a multi-scale parameter controlled thinning method. Finally, we incorporate the initial skeleton into the \(L_1\)-median optimization, and develop a distance field guided \(L_1\)-median to effectively extract the complete skeleton from the point cloud. Our method exhibits the advantages of both the distance field based skeleton extraction methods and the \(L_1\)-median skeleton extraction methods. Our skeleton extraction system is robust and effective, and can be applied to the raw scanned point cloud data.

Plant phenotypic analysis is of great importance to the development of agricultural engineering, and is one of the core issues in crop science and plant breeding. Since plant growth is spatio-temporal and synchronous, understanding the growth and development of individual plants can help to reveal the growth potential of the whole plot and thus improve planting methods. In recent years, the technical means to analyze the growth situation using 3D point cloud data has received extensive attention. The plant point cloud obtained by scanning plants with LiDAR has the characteristics of high resolution, high precision, etc. Periodic scanning of the same plant for spatio-temporal point cloud data sets allows monitoring of growth through subtle changes of plant organs. Organ tracking of growing plants remains challenging due to the lightward nature of growth, the potential for topological changes and the unpredictability of plant growth over time, with the possibility of new leaf growth and leaf death. This paper designs a plant organ growth tracking method based on point cloud. First of all, for growing plants, we have established a crop point cloud spatio-temporal dataset based on two publicly available point cloud datasets. The data set includes four species, tomato, tobacco, sorghum and maize, each species contains complete organ instance labels, and each organ of the same plant has a unique label, which means that the labels of the same organ of an individual at different scan dates correspond one-to-one. Second, this paper proposes a point cloud data-based plant organ growth tracking method, which uses a cost correlation matrix to automatically track growing plant organs. Finally, based on the set of quantitative evaluation metrics, our algorithm achieves a matching accuracy of 82.89% on the plant spatio-temporal dataset and good growth tracking results in the qualitative analysis.