Mobile Mapping Research Articles

Decarbonization and climate change analysis of tunnels in an Asset Management framework through MIRET

The planning and the management of existing infrastructures (and tunnels) is already a central challenge for industrialized countries to manage heritage and strategic infrastructures ensuring a resilient asset against extreme events such as climate change. The functions and tools for assessing and mitigating the risks of climate change are fundamental for identifying and managing the elements of the territory, among which infrastructures play a key role due to their role and the numerous strategic interferences with the environment.To meet this challenge in infrastructures and major civil works like tunnels, an approach for decision-making is being developed in a digital and multidisciplinary environment, which takes the name of MIRET (Management and Identification of the Risk for Existing Tunnels). ETS has established itself in recent years with numerous research and innovations in the field of sustainability and modelling the risk analysis of infrastructures. The risk process is built through a Stakeholders Engagement process involving Owners, Contractors, Designers, Government, Communities and Suppliers, and guided by the team of specialists of the Center for Climate Change (C3) of Rovira I Virgili University.The paper focuses on the MIRET technologies and innovations to reach a more sustainable Asset Management of tunnels. The decarbonization is analyzed for the survey and inspections phase, comparing consolidated multi-dimensional mobile mapping systems and digitalized processes to the relevant baseline. In this framework, a focus is dedicated to two topics developed by ETS: the effort in implementing Artificial Intelligence algorithms for the transition towards Smart Data, and the quantitive approach to assess the climate impacts. The implementations are contextualised on actual case studies involving strategic tunnels in Italy.

Velocity Estimation Using Time-Differenced Carrier Phase and Doppler Shift with Different Grades of Devices: From Smartphones to Professional Receivers

Velocity estimation has a key role in several applications; for instance, velocity estimation in navigation or in mobile mapping systems and GNSSs is currently a common way to achieve reliable and accurate velocity. Two approaches are mainly used to obtain velocity based on GNSS measurements, i.e., Doppler observations and carrier phases differenced in time (that is, TDCP). In a benign environment, Doppler-based velocity can be estimated accurately to within a few cm/s, while TDCP-based velocity can be estimated accurately to within a few mm/s. On the other hand, the TDCP technique is more prone to availability shortage and the presence of blunders. In this work, the two mentioned approaches are tested, using three devices of different grades: a high-grade geodetic receiver, a high-sensitivity receiver, and a GNSS chip mounted on a smartphone. The measurements of geodetic receivers are inherently cleaner, providing an accurate solution, while the remaining two receivers provide worse results. The case of smartphone GNSS chips can be particularly critical owing to the equipped antenna, which makes the measurements noisy and largely affected by blunders. The GNSSs are considered separately in order to assess the performance of the single systems. The analysis carried out in this research confirms the previous considerations about receiver grades and processing techniques. Additionally, the obtained results highlight the necessity of adopting a diagnostic approach to the measurements, such as RAIM-FDE, especially for low-grade receivers.

Accuracy comparison of mobile mapping system for road inventory

Mobile Mapping Systems (MMSs) stand out as the preferred solution for achieving highly precise 3D environmental models, particularly in urban planning, highway mapping, asset inventory, corridor mapping, traffic safety evaluation, autonomous vehicle, digital twin, and emergency response mapping where traditional aerial or satellite surveys often fall short in providing precise data. Understanding the intricate factors that impact the accuracy of mobile mapping is pivotal to harnessing the full potential of this advanced measurement technique. This study analyzes the spatial accuracy of geographical data produced by the mobile mapping method, considering factors such as the speed of the mobile mapping tool, measurement time difference, camera shooting distance of the produced data, and differences in picture shooting distances. The acquired results were examined for their applicability in the production of inventory along the highway route, revealing their practical usability through analysis and findings. This investigation delves into the proficiency and precision benchmarks of mobile mapping systems, specifically in the context of creating road inventory and supporting decision-making for road systems. The study discusses the usability and accuracy criteria of mobile mapping systems for creating transportation inventory and decision support systems.

Preface: Workshop “Robotics for Mapping – SLAM approaches for mobile mapping and robot intelligence”

Preface: Workshop “Robotics for Mapping – SLAM approaches for mobile mapping and robot intelligence”

Preface: Workshop “Robotics for Mapping – SLAM approaches for mobile mapping and robot intelligence”

Preface: Workshop “Robotics for Mapping – SLAM approaches for mobile mapping and robot intelligence”

COMPARATIVE ANALYSIS OF MORPHOLOGICAL (MCSS) AND LEARNING-BASED (SPG) STRATEGIES FOR DETECTING SIGNAGE OCCLUSIONS ALONG TRANSPORTATION CORRIDORS

Abstract. Signage visibility along transportation corridors is critical for drivers in terms of road safety, traffic flow, and enforcement. Traffic signs that are easy to recognize by drivers and autonomous vehicles can help in avoiding accidents and improve safety. Nowadays, Mobile Mapping Systems (MMS) equipped with LiDAR units can scan road network components and its surrounding environment at a normal driving speed while collecting accurate geospatial data. Most traffic signs have well-defined geometric characteristics (e.g., linear or planar features) which can be identified in the 3D LiDAR data acquired by MMS. Therefore, MMS LiDAR data are an ideal source to recognize traffic signs. In addition to traffic sign detection, MMS can also identify vegetation along the right-of-way and evaluate signage visibility. Thus, this paper presents a framework for using MMS LiDAR data for traffic sign and vegetation detection which is a prerequisite for signage visibility analysis. For signage and vegetation detection, two alternative strategies are adopted: 1) a morphological approach and 2) a learning-based approach. For the geometric/morphological approach, Multi-Class Simultaneous Segmentation (MCSS) is utilized in this study. As for the learning-based strategy, semantic segmentation of LiDAR data are performed using Super Point Graph (SPG). Lastly, signage visibility analysis is conducted based on the occlusion rate assessed from different driver’s viewpoints.

MFSCNN: APPENDING A MASKED BRANCH TO FAST-SCNN TO IMPROVE ROAD MARKING EXTRACTION ON SPARSE MLS POINT CLOUD-DERIVED IMAGES

Abstract. With the rise of self-driving cars, an increasing number of vehicles are equipped with low-cost light detection and ranging (LiDAR) sensors that could potentially serve as a massive mobile mapping resource, particularly for jobs that require multiple and frequent scanning, such as maintaining dynamic high-definition maps or digital twins. However, low-cost LiDAR sensors produce sparser point clouds during scanning which can make deep learning techniques for the automatic retrieval of features difficult like extracting road markings. In this work, we aim to improve the performance of a convolutional neural network (CNN) model for road marking extraction from sparse mobile LiDAR scanning (MLS) point cloud-derived images. We propose the modification of the Fast-SCNN model structure by adding a 2D convolution branch with masking in the feature fusion step: MFSCNN. To retain speed we only use MFSCNN to boost model training and still utilize Fast-SCNN for inference. Our results indicate potential, with a 4.6% increase in mean f1-score and an 8% decrease in uncertainty for the road marking class after multiple trials. Additionally, this research aims to support and increase research interest in lower-cost LiDARs for mobile mapping.

OBJECT DETECTION AND LOCALISATION FOR BIM ENRICHMENT

Abstract. The use of Building Information Models (BIM) during the entire life cycle of a building or facility requires an up-to-date and detailed digital representation. Nowadays the BIM focus mostly lies on the design and construction phase of the building and is rarely used for the rest of the life cycle. For further use the BIM must be updated after construction and enriched with appliance objects such as safety equipment, heating elements to enhance its usability. In this work a new approach is presented to detect and locate appliance objects in a three dimensional environment using object detection in images created by indoor mobile mapping devices. By detecting and locating these appliance objects in the three dimensional world they can be included in the BIM model. This approach enables the detection and localization needed for the placement of placeholders or detailed geometric models of those appliances in the BIM model or other digital representations. Resulting in an increased level of detail and usability of the digital representation of the building or facility during its further life cycle. Overall, this work demonstrates and examines the accuracy of the detection and localization of appliance objects.

STRATEGY ON HIGH-DEFINITION POINT CLOUD MAP CREATION FOR AUTONOMOUS DRIVING IN HIGHWAY ENVIRONMENTS

Abstract. In recent years, a lot of researchers have been trying the development of efficient ways to create HD maps with centimeter-level precision. Mobile mapping system (MMS) produce 3D HD point cloud map of the surrounding by integrating navigation (i.e., direct georeferencing or DG process) and high-resolution imaging sensor data. Unfortunately, in partially environments, the provided accuracy of the GNSS system degrades dramatically. In order to constraint the drift and correct the georeferenced point cloud map, ground control points (GCPs) are placed along the road. Moreover, there are approaches which use laser-based point cloud registration techniques to construct the point cloud map. However, all promising mapping techniques which currently use the laser as the core sensor for mapping the high-definition point cloud map may not be promised to construct the point cloud map in partially or unfriendly environments. As a literature review and result, a suitable approach for creating the promising point cloud map is to combine the INS/GNSS navigation solution, LiDAR matching techniques, and GCPs. Thus, this study introduces the HD point cloud map generation method that can potentially help researchers create personalized and globalized HD point cloud maps and develop new HD point cloud map generation methodologies.

3D MODELING OF ROAD INFRASTRUCTURES ACCORDING TO CITYGML 3.0 AND ITS CITYJSON ENCODING

Abstract. This study proposes a systematic approach for standardized 3D modeling of road infrastructures based on CityGML 3.0 and its CityJSON encoding. The approach involves generating a LoD2 3D model of the road infrastructure based on a semi-automatic extraction of linear features from mobile mapping LiDAR data. This enables geometric and semantic modeling of the roads. A codification system is proposed to assign predefined codes to each linear feature, allowing each section and intersection and each road surface to be modeled separately. The resulting model is converted to a CityJSON file, stored in a document-oriented database and visualized through a web application. The proposed approach provides a cost-effective alternative to traditional manual modeling methods while maintaining a high level of accuracy and consistency. It also considers the validation of data schema and geometric primitives to ensure that any non-conformity with CityJSON schemas, and any topological and/or geometric errors can be detected and then corrected. This is important since schema changes in new versions of CityJSON can result in compatibility issues, while geometric and topological errors can affect the accuracy of 3D models and ultimately lead to inaccurate simulation outcomes or analysis results.

A study on the aptitude of color hue, value, and transparency for geographic relevance encoding in mobile maps

ABSTRACT Depicting the right amount of information on mobile maps is still challenging. With this research we investigate which of the three visual variables hue, value, and transparency optimally directs visual attention to the correct encoding of the geographic relevance of on small displays? We hypothesize that transparency is best suited to let users see symbols in the correct order of high to low relevance and that they intuitively understand this encoding of symbols for value and transparency, but not for hue. In a user study with 27 participants, we tested the suitability of the three visual variables for mobile maps with eye tracking. Results for a perceptual viewing task where no further knowledge or a task was provided, showed no significant differences between the three visual variables for time to first fixation and scanning the symbols in the correct order of relevance. A second search task for the most relevant restaurant, results showed that only transparency seems to reliably encode geographic relevance. “Darker is more” is only partially understood and the traffic light metaphor is unsuitable for encoding ordered geographic relevance on mobile maps. These results are in line with other work and offer guidelines for mobile map design.

TOWARDS ACCURATE INSTANCE SEGMENTATION IN LARGE-SCALE LIDAR POINT CLOUDS

Abstract. Panoptic segmentation is the combination of semantic and instance segmentation: assign the points in a 3D point cloud to semantic categories and partition them into distinct object instances. It has many obvious applications for outdoor scene understanding, from city mapping to forest management. Existing methods struggle to segment nearby instances of the same semantic category, like adjacent pieces of street furniture or neighbouring trees, which limits their usability for inventory- or management-type applications that rely on object instances. This study explores the steps of the panoptic segmentation pipeline concerned with clustering points into object instances, with the goal to alleviate that bottleneck. We find that a carefully designed clustering strategy, which leverages multiple types of learned point embeddings, significantly improves instance segmentation. Experiments on the NPM3D urban mobile mapping dataset and the FOR-instance forest dataset demonstrate the effectiveness and versatility of the proposed strategy.

GROUND FILTERING OF CO-REGISTERED MOBILE AND STATIONARY LASER SCANS BY USING SUPERPOINTS IN RANSAC PLANES

Abstract. Ground filtering is an important tool for many applications. The high variability of landscapes makes it necessary to perform its computation with 3D points as the only input, that is, with as few as possible algorithm parameters and without any training data. In the case of terrestrial laser scans, an additional challenge comes from a highly inhomogeneous point density. The SiRP algorithm on ground filtering, relying on intermediate validating superpixels as ground or non-ground, was previously developed for airborne point clouds. We consider a dataset acquired by a mobile mapping system mounted on a car, which was extended by additional stationary laser scans. The registration algorithm is based on hierarchical merging. Afterwards, SiRP was applied to both uni-modal and multi-modal point clouds. Impressive qualitative and quantitative classification results with around 97 % on overall accuracy were obtained and discussed.

BACKPACK SYSTEM FOR CAPTURING 3D POINT CLOUDS OF FORESTS

Abstract. A 3D model can be useful for inventory management and monitoring of forests. For this task, we present our prototype mobile mapping backpack system for collecting 3D point clouds of forest vegetation. However, data collection in forests is challenging due to unreliable GNSS positioning, moving objects caused by winds, unclear object edges, and uneven ground to walk on. We include LiDAR and IMU with Simultaneous Localisation and Mapping (SLAM). We describe in detail our backpack system, its development for the forest environment, and evaluate it. Furthermore, we compare two open code SLAM algorithms for ROS, as well as data collection and laser scan quality between TLS and MLS for forest environments. Finally, we installed and tested the MicaSense multi-spectral camera on our backpack and discuss the advantages and drawbacks of it. We conclude, that the backpack is convenient to use in forest environments and produces a good point cloud. It can be carried easily off trail and on rough terrain. The system needs less storage space, computation and less collection time than TLS.

ESTABLISHMENT OF HD MAPS VERIFICATION AND VALIDATION PROCEDURE WITH OPENDRIVE AND AUTOWARE (LANELET2) FORMATS

Abstract. Mobile mapping technologies, for example multi-sensor integration and multi-platform mapping technology, have developed and improved over recent decades, various applications such as conventional mapping scenarios, rapid disaster response, smart city, and autonomous vehicle application arise synchronously. Especially, autonomous driving vehicles have made enormous progress. High-definition (HD) maps are key for autonomous driving because of their high accuracy and rich information of road scenes. However, how to make sure that HD maps are suitable for autonomous vehicle requirement is an important topic. The HD maps guidelines and standards in Taiwan are released since 2018 and mainly focus on point cloud and shape file format. In this paper, a procedure for the verification and validation of HD maps for OpenDRIVE and Autoware (Lanelet2) is proposed. It discusses about the verification strategies, suggestion review item, recommendation tools, and process. As shown by our preliminary results, the proposed process can conform not only in closed area but also public road. These issues can help reducing HD maps production costs. When the foundation of HD maps accomplishes, the autonomous driving techniques can naturally complement. The vision of full automation vehicle will come true rapidly in the future.

Collective interactions, collaborative inhibition, and shared spatial knowledge

ABSTRACT Research on spatial mental representations focuses on individual mental maps and spatial knowledge. This exploratory study investigates instead collective interactions, collaborative memory, and the sharing of spatial knowledge. Based on the principle of collaborative inhibition (i.e., people recall information less effectively in groups), we posed the following research question: How do collective interactions, occurring during environmental exploration and group drawing sessions, affect collaborative inhibition, and the quality of sketch maps designed collectively? We conducted in situ explorations in Plaine St-Denis (France) with real-time tracking, followed by individual and group drawing sessions. This experiment involved 118 participants divided into three groups: (1) solo explorations without devices; (2) solo explorations with a mobile mapping application; (3) collective explorations using the same application enhanced with interaction features (viewing collective routes and photos of visited places). The comparison of the total number of entities found on individual mental maps with those included in collective sketch maps reveals that collaborative inhibition applies to spatial memory. Additional findings indicate that the use of a map, combined with collective interactions, mitigates collaborative inhibition and increases the accuracy of the sketch maps. However, the effect of such interactions on group dynamics remains unclear as of now.

Low-Cost Webcam Camera and GNSS Integration for Updating Home Data using AI Principles

PDAM Surya Sembada Surabaya City responds to Surabaya's water service demand by pioneering a cost-effective mobile mapping tech. This innovation integrates 3 MP webcams, GNSS, and IMU sensors to update customer building data efficiently. Precise photo data with location and movement info within 1-meter proximity captures accurate building coordinates. Applying georeferencing and Cosine Similarity reduces redundant data by 50%. GNSS, IMU sensors, and VGG-16 model aid data collection and analysis. This comprehensive approach optimizes PDAM Surabaya's operations, ensuring timely and accurate customer updates amid Surabaya's population growth, promising enhanced service delivery while managing costs.

Users’ Visual Experience During Temporal Navigation in Forecast Weather Maps on Mobile Devices

This study investigated the impact of graphical user interface (GUI) design on the efficiency and effectiveness of map-based tasks on mobile devices, using time-based weather data as a case study. Three different GUI designs (button-type, circle-type, and slidebar) were tested in a between-subjects design, with 50 participants completing a set of map-based tasks on each GUI design. The results showed that GUI design significantly affected the effectiveness of map-based tasks. Participants performed better at tasks involving the search for the highest and lowest temperature amplitudes on the button-type GUI whereas the circle-type GUI showed lower effectiveness for tasks involving the search for day temperatures. Analysis of the visual attention distribution based on fixation count revealed that different GUI designs led to different patterns of visual attention. The study highlights the importance of considering GUI design in the development of mobile map applications, particularly for map-based tasks involving time-based data. The study shows that separating the date from the time navigation panel reduces necessary visual focus on the GUI itself and is a valuable insight for future GUI design.

Generating lane-level road networks from high-precision trajectory data with lane-changing behavior analysis

– Recent advances in mobile mapping systems have facilitated the collection of high-precision trajectory data in centimeter positioning accuracy. It provides the potential to infer lane-level road networks, which are essential for autonomous driving navigation. This task is challenging due to the complicated lane merging and diverging structures as well as the lane-changing patterns in trajectory data. This paper presents a lane-level road network generation method from high-precision trajectory data with lane-changing behavior analysis. Trajectories are firstly partitioned by detecting road intersections and changes in lane structure. Subsequently, in regions with consistent lane structure, a principal curve fitting algorithm is developed to extract lane centerlines. Erroneous lanes generated by lane-changing behavior are pruned based on a constructed lane intersection graph. In regions with merging and diverging lanes, a lane-group fitting algorithm is designed. This algorithm estimates lane locations by incorporating a Gaussian mixture model with lane width prior knowledge and then infers lane-level topological structures using trajectory flow information. The proposed method is evaluated on a real-world high-precision trajectory dataset. Comprehensive experiments demonstrate that it outperforms state-of-the-art methods in four metrics. Under complex scenarios, the method is capable of generating lane-level road networks with higher completeness and fewer fragments.

Tree species classification on images from airborne mobile mapping using ML.NET

ABSTRACT Deep learning is a powerful tool for automating the process of recognizing and classifying objects in images. In this study, we used ML.NET, a popular open-source machine learning framework, to develop a model for identifying tree species in images obtained from airborne mobile mapping. These high-resolution images can be used to create detailed maps of the landscape. They can also be analyzed and processed to extract information about visual features, including tree species recognition. The deep learning model was trained using ML.NET to classify two tree species based on the combination of airborne mobile mapping images. Our approach yielded impressive results, with a maximum classification accuracy of 93.9%. This demonstrates the effectiveness of combining imagery sources with deep learning tools in ML.NET for efficient and accurate tree species classification. This study highlights the potential of the ML.NET framework for automating object classification and can provide valuable insights and information for forestry management and conservation efforts. The primary objective of this research was to evaluate the effectiveness of an approach for identifying tree species through a model generated using a combination of ortho and oblique images captured by a mobile mapping system.