The precise extraction of road boundaries is an essential task to obtain road infrastructure data that can support various applications, such as maintenance, autonomous driving, vehicle navigation, and the generation of high-definition maps (HD map). Despite promising outcomes in prior studies, challenges persist in road extraction, particularly in discerning diverse road types. The proposed methodology integrates state-of-the-art techniques like DBSCAN and RANSAC, aiming to establish a universally applicable approach for diverse mobile mapping systems. This effort represents a pioneering step in extracting road information from image-based point cloud data. To assess the efficacy of the proposed method, we conducted experiments using a large-scale dataset acquired by two mobile mapping systems on the Yıldız Technical University campus; one system was configured as a mobile LiDAR system (MLS), while the other was equipped with cameras to operate as a photogrammetry-based mobile mapping system (MMS). Using manually measured reference road boundary data, we evaluated the completeness, correctness, and quality parameters of the road extraction performance of our proposed method based on two datasets. The completeness rates were 93.2% and 84.5%, while the correctness rates were 98.6% and 93.6%, respectively. The overall quality of the road curb extraction was 93.9% and 84.5% for the two datasets. Our proposed algorithm is capable of accurately extracting straight or curved road boundaries and curbs from complex point cloud data that includes vehicles, pedestrians, and other obstacles in urban environment. Furthermore, our experiments demonstrate that the algorithm can be applied to point cloud data acquired from different systems, such as MLS and MMS, with varying spatial resolutions and accuracy levels.

The extraction of artificial and natural features using light detection and ranging (Lidar) data is a fundamental task in many fields of research for environmental science. In this study, the possibility of using self-organising maps (SOM), which is an unsupervised artificial neural network classification method to extract the bare earth surface and features from airborne Lidar data, was investigated for two different urban areas. The effect of the enlargement of the study area was analysed using the proposed approach. The appropriate weights of SOM inputs, which are 3D coordinates and intensity, obtained from a Lidar point cloud were determined by using Pearson's chi-squared independence test. The weighted SOM feature extraction performance was better than that of the unweighted SOM. The filtering results of SOM to separate ground and non-ground data were also compared with those obtained by the adaptive TIN filtering algorithm. Most of the non-ground features could be removed by the weighted SOM.

Unmanned Aerial Systems (UAS) equipped with LiDAR sensors are increasingly used for topographic surveying and mapping due to their ability to generate high-quality point clouds and penetrate vegetation. While photogrammetry remains a common approach in UAS mapping, its limitations in vegetated environments—where it often struggles to accurately capture ground surface details—have led to the adoption of LiDAR. LiDAR, with its capability to penetrate vegetation, provides more precise terrain observations and detailed representations of the underlying ground surface, even in densely vegetated areas. This study compares the performance of two LiDAR systems: the RESEPI Gen-I-M2X sensor from Inertial Labs, mounted on a WingtraOne Gen II fixed-wing UAS using post-processed kinematic (PPK) corrections; and the Zenmuse L2 sensor integrated with the DJI Matrice 350 RTK, utilizing real-time kinematic (RTK) corrections with additional PPK adjustments based on observation files from a local Continuously Operating Reference Station (CORS) acting as the base. Both platforms were flown over the same area, with point clouds analyzed across three distinct conditions: open terrain, urban development, and wooded areas. A total of 135 GNSS-measured reference points were deployed, with 10 designated as ground control points (GCPs) to enhance vertical accuracy, while the remaining 125 points served as checkpoints for validation. Some checkpoints were located at the center of manhole covers, others at painted arrow markers on roadways, but the majority—especially those in wooded areas—were natural points without signalization. The Zenmuse L2 datasets were processed in DJI Terra, generating point clouds with and without GCP integration. In contrast, the RESEPI Gen-I-M2X datasets relied solely on PPK corrections, as the processing software does not support GCP integration. This study evaluates the accuracy and noise levels of the point clouds in varying environments, focusing on terrain representation by the LiDAR sensors. The findings provide insights into the strengths and limitations of each platform and correction strategy, offering guidance for selecting appropriate UAS LiDAR systems for specific surveying and mapping applications. This research contributes to the growing body of work on UAS LiDAR by highlighting key factors that influence data quality, including sensor selection, correction methods, and environmental conditions.

The Smart City concept is taking momentum recently as big metropolises as well as mid-size cities are intensifying their efforts to improve the life of people living in dense urban environment. Local governments are eager to have up-to-date information of every aspect of city life, including environmental data, such as air and water quality parameters; mobility data, such as traffic flow, including vehicles, transit passengers; crowd control, such as public events, mobility in hospitals; life quality data, such as social status, education level, health records; etc. Monitoring all these very different data streams in space and time is a formidable challenge. While on the data acquisition side, tremendous progress has been achieved, as sensors have been deployed in increasingly large numbers on both mobile and static platforms, there is a lack of creating accurate geotags, as the quality of georeferencing varies over a large scale. It is important to note that the data acquisition is becoming largely customer-based, as smart devices are efficient sensor systems and with advancing communication technologies, crowdsourcing is quickly becoming the dominant data source on mobile platforms. In this paper, we investigate the potential to exploit the ranging capabilities of imaging and communication sensors and use the strength of the spatial network formed by the sensors to improve the georeferencing of a group of platforms operating in a close environment, such as UAS swarm or a platoon of autonomous vehicles. Transportation in cities and in general mobility are of great interest to Smart Cities, they represent one of the most significant components of the activities, so having an optimized transportation system is essential to reduce carbon footprint, decrease commute time, and just improve the quality of life. To assess the performance of collaborative navigation based accurate georeferencing, data was acquired at a simulated intersection area at The Ohio State University, where multiple vehicles, pedestrians and cyclists were moving around. In addition, drones were flying above the area. Here we report about our initial results.