PCEDNet: A Lightweight Neural Network for Fast and Interactive Edge Detection in 3D Point Clouds

In recent years, Convolutional Neural Networks (CNN) have proven to be efficient analysis tools for processing point clouds, e.g., for reconstruction, segmentation, and classification. In this article, we focus on the classification of edges in point clouds, where both edges and their surrounding are described. We propose a new parameterization adding to each point a set of differential information on its surrounding shape reconstructed at different scales. These parameters, stored in a Scale-Space Matrix (SSM) , provide a well-suited information from which an adequate neural network can learn the description of edges and use it to efficiently detect them in acquired point clouds. After successfully applying a multi-scale CNN on SSMs for the efficient classification of edges and their neighborhood, we propose a new lightweight neural network architecture outperforming the CNN in learning time, processing time, and classification capabilities. Our architecture is compact, requires small learning sets, is very fast to train, and classifies millions of points in seconds.

GeoAI: a review of Artificial Intelligence approaches for the interpretation of complex Geomatics data

Researchers have explored the benefits and applications of modern artificial intelligence (AI) algorithms in different scenario. For the processing of geomatics data, AI offers overwhelming opportunities. Fundamental questions include how AI can be specifically applied to or must be specifically created for geomatics data. This change is also having a significant impact on geospatial data. The integration of AI approaches in geomatics has developed into the concept of Geospatial Artificial Intelligence (GeoAI), which is a new paradigm for geographic knowledge discovery and beyond. However, little systematic work currently exists on how researchers have applied AI for geospatial domains. Hence, this contribution outlines AI-based techniques for analysing and interpreting complex geomatics data. Our analysis has covered several gaps, for instance defining relationships between AI-based approaches and geomatics data. First, technologies and tools used for data acquisition are outlined, with a particular focus on RGB images, thermal images, 3D point clouds, trajectories, and hyperspectral/multispectral images. Then, how AI approaches have been exploited for the interpretation of geomatic data is explained. Finally, a broad set of examples of applications are given, together with the specific method applied. Limitations point towards unexplored areas for future investigations, serving as useful guidelines for future research directions.

Accuracy Assessment of 3D Model Based on Laser Scan and Photogrammetry Data

A three-dimensional (3D) model extraction represents the best way to reflect the reality in all details. This explains the trends and tendency of many scientific disciplines towards making measurements, calculations and monitoring in various fields using such model. Although there are many ways to produce the 3D model like as images, integration techniques, and laser scanning, however, the quality of their products is not the same in terms of accuracy and detail. This article aims to assess the 3D point clouds model accuracy results from close range images and laser scan data based on Agi soft photoscan and cloud compare software to determine the compatibility of both datasets for several applications. College of Science, Departments of Mathematics and Computer in the University of Baghdad campus were exploited to create the proposed 3D model as this area location, which is one of the distinctive features of the university, allows making measurements freely from all sides. Results of this study supported by statistical analysis including 2 sample T-test and RMSE calculation in addition to visual comparison. Through this research, we note that the laser3D model provides many points in a short time, so it will reduce the field work and also its data is faster in processing to produce a reliable model of the scanned area compared with data derived from photogrammetry, then the difference were computed for all the reference points.

Potential Use of Drone Ultra-High-Definition Videos for Detailed 3D City Modeling

Ongoing developments in video resolution either using consumer-grade or professional cameras has opened opportunities for different applications such as in sports events broadcasting and digital cinematography. In the field of geoinformation science and photogrammetry, image-based 3D city modeling is expected to benefit from this technology development. Highly detailed 3D point clouds with low noise are expected to be produced when using ultra high definition UHD videos (e.g., 4K, 8K). Furthermore, a greater benefit is expected when the UHD videos are captured from the air by consumer-grade or professional drones. To the best of our knowledge, no studies have been published to quantify the expected outputs when using UHD cameras in terms of 3D modeling and point cloud density. In this paper, a quantification is shown about the expected point clouds and orthophotos qualities when using UHD videos from consumer-grade drones and a review of which applications they can be applied in. The results show that an improvement in 3D models of ≅65% relative accuracy and ≅90% in point density can be attained when using 8K video frames compared with HD video frames which will open a wide range of applications and business cases in the near future.

Development of a Low-Cost System for 3D Orchard Mapping Integrating UGV and LiDAR

Growing evaluation in the early stages of crop development can be critical to eventual yield. Point clouds have been used for this purpose in tasks such as detection, characterization, phenotyping, and prediction on different crops with terrestrial mapping platforms based on laser scanning. 3D model generation requires the use of specialized measurement equipment, which limits access to this technology because of their complex and high cost, both hardware elements and data processing software. An unmanned 3D reconstruction mapping system of orchards or small crops has been developed to support the determination of morphological indices, allowing the individual calculation of the height and radius of the canopy of the trees to monitor plant growth. This paper presents the details on each development stage of a low-cost mapping system which integrates an Unmanned Ground Vehicle UGV and a 2D LiDAR to generate 3D point clouds. The sensing system for the data collection was developed from the design in mechanical, electronic, control, and software layers. The validation test was carried out on a citrus crop section by a comparison of distance and canopy height values obtained from our generated point cloud concerning the reference values obtained with a photogrammetry method. A 3D crop map was generated to provide a graphical view of the density of tree canopies in different sections which led to the determination of individual plant characteristics using a Python-assisted tool. Field evaluation results showed plant individual tree height and crown diameter with a root mean square error of around 30.8 and 45.7 cm between point cloud data and reference values.

Learning Rotations

Many problems in computer vision today are solved via deep learning. Tasks like pose estimation from images, pose estimation from point clouds or structure from motion can all be formulated as a regression on rotations. However, there is no unique way of parametrizing rotations mathematically: matrices, quaternions, axis-angle representation or Euler angles are all commonly used in the field. Some of them, however, present intrinsic limitations, including discontinuities, gimbal lock or antipodal symmetry. These limitations may make the learning of rotations via neural networks a challenging problem, potentially introducing large errors. Following recent literature, we propose three case studies: a sanity check, a pose estimation from 3D point clouds and an inverse kinematic problem. We do so by employing a full geometric algebra (GA) description of rotations. We compare the GA formulation with a 6D continuous representation previously presented in the literature in terms of regression error and reconstruction accuracy. We empirically demonstrate that parametrizing rotations as bivectors outperforms the 6D representation. The GA approach overcomes the continuity issue of representations as the 6D representation does, but it also needs fewer parameters to be learned and offers an enhanced robustness to noise. GA hence provides a broader framework for describing rotations in a simple and compact way that is suitable for regression tasks via deep learning, showing high regression accuracy and good generalizability in realistic high-noise scenarios.

Clustering-Based Plane Segmentation Neural Network for Urban Scene Modeling

Urban scene modeling is a challenging but essential task for various applications, such as 3D map generation, city digitization, and AR/VR/metaverse applications. To model man-made structures, such as roads and buildings, which are the major components in general urban scenes, we present a clustering-based plane segmentation neural network using 3D point clouds, called hybrid K-means plane segmentation (HKPS). The proposed method segments unorganized 3D point clouds into planes by training the neural network to estimate the appropriate number of planes in the point cloud based on hybrid K-means clustering. We consider both the Euclidean distance and cosine distance to cluster nearby points in the same direction for better plane segmentation results. Our network does not require any labeled information for training. We evaluated the proposed method using the Virtual KITTI dataset and showed that our method outperforms conventional methods in plane segmentation. Our code is publicly available.

ForestScanner: A mobile application for measuring and mapping trees with LiDAR-equipped iPhone and iPad

Terrestrial laser scanning (TLS) is becoming increasingly popular as an alternative means to conventional forest inventory methods. By gauging the distances to multiple points on the surrounding object surfaces, TLS acquires 3D point clouds from which tree sizes and spatial distributions can be rapidly estimated. However, the high cost and specialized skills required for TLS have put it out of reach for many potential users. We here introduce ForestScanner, a free, mobile application that allows TLS-based forest inventories by means of iPhone or iPad with a built-in LiDAR sensor. ForestScanner does not require any manual analysis of 3D point clouds. As the user scans trees with an iPhone/iPad, ForestScanner estimates the stem diameters and spatial coordinates based on real-time instance segmentation and circle fitting. The users can visualize, check, and share the results of scanning in situ. By using ForestScanner, we measured the stem diameters and spatial coordinates of 672 trees within a 1 ha plot in 1 h 39 min with an iPhone and in 1 h 38 min with an iPad (diameter ≥5 cm; detection rate = 100%). ForestScanner reduced the person-hours required for measuring diameters to 25.7%, mapping trees to 9.3%, and doing both to 6.8% of the person-hours taken using a dimeter tape and a conventional surveying method. The diameters measured by ForestScanner and diameter tape were in good agreement; R2=0.963 for iPhone and R2=0.961 for iPad. ForestScanner and the conventional surveying system showed almost identical results for tree mapping (assessed by the spatial distances among trees within 0.04 ha subplots); Mantel R2=0.999 for both iPhone and iPad. Our results indicate that ForestScanner enables cost-, labor-, and time-efficient forest inventories. The application can increase the accessibility to TLS for people beyond specialists and enhance resource assessments and biodiversity monitoring in forests worldwide.