Tree heights are the principal variables for forest plantation inventory. The increasing availability of high-resolution three-dimensional (3D) point clouds derived from low-cost Unmanned Aerial Vehicle (UAV) and modern photogrammetry offers an opportunity to generate a Canopy Height Model (CHM) in the mountainous areas. In this paper, we assessed the capabilities of tree height estimation using UAV-based Structure-from-Motion (SfM) photogrammetry and Semi-Global Matching (SGM). The former is utilized to generate 3D geometry, while the latter is used to generate dense point clouds from UAV imagery. The two algorithms were coupled with a Radial Basis Function (RBF) neural network to acquire CHMs in mountainous areas. This study focused on the performance of Digital Terrain Model (DTM) interpolation over complex terrains. With the UAV-based image acquisition and image-derived point clouds, we constructed a 5 cm-resolution Digital Surface Model (DSM), which was assessed against 14 independent checkpoints measured by a Real-Time Kinematic Global Positioning System RTK GPS. Results showed that the Root Mean Square Errors (RMSEs) of horizontal and vertical accuracies are approximately 5 cm and 10 cm, respectively. Bare-earth Index (BEI) and Shadow Index (SI) were used to separate ground points from the image-derived point clouds. The RBF neural network coupled with the Difference of Gaussian (DoG) was exploited to provide a favorable generalization for the DTM from 3D ground points with noisy data. CHMs were generated using the height value in each pixel of the DSM and by subtracting the corresponding DTM value. Individual tree heights were estimated using local maxima algorithm under a contour-surround constraint. Two forest plantations in mountainous areas were selected to evaluate the accuracy of estimating tree heights, rather than field measurements. Results indicated that the proposed method can construct a highly accurate DTM and effectively remove nontreetop maxima. Furthermore, the proposed method has been confirmed to be acceptable for tree height estimation in mountainous areas given the strong linear correlation of the measured and estimated tree heights and the acceptable t-test values. Overall, the low-cost UAV-based photogrammetry and RBF neural network can yield a highly accurate DTM over mountainous terrain, thereby making them particularly suitable for rapid and cost-effective estimation of tree heights of forest plantation in mountainous areas.
Automatic building extraction using a single data type, either 2D remotely-sensed images or light detection and ranging 3D point clouds, remains insufficient to accurately delineate building outlines for automatic mapping, despite active research in this area and the significant progress which has been achieved in the past decade. This paper presents an effective approach to extracting buildings from Unmanned Aerial Vehicle (UAV) images through the incorporation of superpixel segmentation and semantic recognition. A framework for building extraction is constructed by jointly using an improved Simple Linear Iterative Clustering (SLIC) algorithm and Multiscale Siamese Convolutional Networks (MSCNs). The SLIC algorithm, improved by additionally imposing a digital surface model for superpixel segmentation, namely 6D-SLIC, is suited for building boundary detection under building and image backgrounds with similar radiometric signatures. The proposed MSCNs, including a feature learning network and a binary decision network, are used to automatically learn a multiscale hierarchical feature representation and detect building objects under various complex backgrounds. In addition, a gamma-transform green leaf index is proposed to truncate vegetation superpixels for further processing to improve the robustness and efficiency of building detection, the Douglas–Peucker algorithm and iterative optimization are used to eliminate jagged details generated from small structures as a result of superpixel segmentation. In the experiments, the UAV datasets, including many buildings in urban and rural areas with irregular shapes and different heights and that are obscured by trees, are collected to evaluate the proposed method. The experimental results based on the qualitative and quantitative measures confirm the effectiveness and high accuracy of the proposed framework relative to the digitized results. The proposed framework performs better than state-of-the-art building extraction methods, given its higher values of recall, precision, and intersection over Union (IoU).
Building extraction from airborne Light Detection and Ranging (LiDAR) point clouds is a significant step in the process of digital urban construction. Although the existing building extraction methods perform well in simple urban environments, when encountering complicated city environments with irregular building shapes or varying building sizes, these methods cannot achieve satisfactory building extraction results. To address these challenges, a building extraction method from airborne LiDAR data based on multi-constraints graph segmentation was proposed in this paper. The proposed method mainly converted point-based building extraction into object-based building extraction through multi-constraints graph segmentation. The initial extracted building points were derived according to the spatial geometric features of different object primitives. Finally, a multi-scale progressive growth optimization method was proposed to recover some omitted building points and improve the completeness of building extraction. The proposed method was tested and validated using three datasets provided by the International Society for Photogrammetry and Remote Sensing (ISPRS). Experimental results show that the proposed method can achieve the best building extraction results. It was also found that no matter the average quality or the average F1 score, the proposed method outperformed ten other investigated building extraction methods.
Most filtering algorithms suffer from complex parameter settings or threshold adjusting. To solve this problem, this paper proposes an improved skewness balancing filtering algorithm based on thin plate spline (TPS) interpolation. The proposed algorithm filters the nonground points in an iterative manner. A reference surface that reflects the fluctuation of the terrain is generated using the TPS interpolation method. Accordingly, the elevation difference from each point to the surface can be calculated. By applying the skewness balancing principle to these elevation differences, nonground points can be removed automatically. To verify the validity and robustness of the proposed method, the datasets provided by the International Society for Photogrammetry and Remote Sensing (ISPRS) were adopted. The experimental results show that this presented method can adapt to complex environments and achieve a higher filtering accuracy than the traditional skewness balancing algorithm. Moreover, in comparison with the other eight filtering methods tested by the ISPRS and four improved filtering methods proposed recently, the proposed method achieved an average total error of 5.39%, which is smaller than that of most of these other methods.
Most 3D CityGML building models in street-view maps (e.g., Google, Baidu) lack texture information, which is generally used to reconstruct real-scene 3D models by photogrammetric techniques, such as unmanned aerial vehicle (UAV) mapping. However, due to its simplified building model and inaccurate location information, the commonly used photogrammetric method using a single data source cannot satisfy the requirement of texture mapping for the CityGML building model. Furthermore, a single data source usually suffers from several problems, such as object occlusion. We proposed a novel approach to achieve CityGML building model texture mapping by multiview coplanar extraction from UAV remotely sensed or terrestrial images to alleviate these problems. We utilized a deep convolutional neural network to filter out object occlusion (e.g., pedestrians, vehicles, and trees) and obtain building-texture distribution. Point-line-based features are extracted to characterize multiview coplanar textures in 2D space under the constraint of a homography matrix, and geometric topology is subsequently conducted to optimize the boundary of textures by using a strategy combining Hough-transform and iterative least-squares methods. Experimental results show that the proposed approach enables texture mapping for building façades to use 2D terrestrial images without the requirement of exterior orientation information; that is, different from the photogrammetric method, a collinear equation is not an essential part to capture texture information. In addition, the proposed approach can significantly eliminate blurred and distorted textures of building models, so it is suitable for automatic and rapid texture updates.
Tree species classification is a ubiquitous task in the forest inventory field. Only directly measured feature vectors have been applied to most existing methods that use LiDAR technology for tree species classification. As a result, it is difficult to obtain a satisfactory tree species classification performance. To solve this challenge, the authors of this paper developed two new kinds of feature vectors, including fractal geometry-based feature vectors and quantitative structural model (QSM)-based feature vectors. In terms of fractal geometry, both two fractal parameters were extracted as feature vectors for reflecting how tree architecture is distributed in three-dimensional space. In terms of QSM, the ratio of length change and the ratio of radius change of different branches were extracted as feature vectors. To reduce the feature vector dimensionality and explore valuable feature vectors, feature vector dimension reduction was conducted using the classification and regression tree (CART). Five hundred and sixty-eight individual trees with five tree species were selected for evaluating the performance of the developed feature vectors. The experimental results indicate that the tree species of Fagus sylvatica achieved the highest overall accuracy, which is 98.06%, while Quercus petraea obtained the lowest overall accuracy, which is 96.65%. Four other classical supervised learning methods were adopted for comparison. The comparison result indicates that the proposed method outperformed the other four supervised learning methods no matter which accuracy indicator was adopted. In comparison with the relevant method, the eight feature vectors developed in this paper also performed much better. This indicates that the fractal geometry-based feature vectors and QSM-based feature vectors developed in this paper can effectively improve the performance of tree species classification.