Earticle

Precise 3D modeling of accident scenes is becoming increasingly important for highly reliable traffic accident reconstruction, especially in complex road environments such as ramps and sloped sections. This study evaluates the geometric accuracy and simulation applicability of 3D road models generated using a consumer-grade 360 camera and photogrammetry technology. The test site was conducted at the circular ramp section of K-City, a dedicated autonomous driving test facility. The point cloud data (PCD) obtained through 360 camera-based photogrammetry was compared with high-precision Mobile Mapping System (MMS) data, which served as the ground truth. To analyze the practical error in a simulation environment, vehicle trajectories (x, y, z coordinates) were extracted from PC-Crash using both models. The study analyzes the similarity between the two trajectories through 3D Euclidean distance, root mean square error (RMSE), and slope change rates. The results provide quantitative evidence on the reliability of low-cost 3D modeling for accident reconstruction and suggest its potential and limitations in modeling complex road geometries.

LiDAR는 자율주행차량이 주변의 객체를 정확하게 식별하고 안전을 강화하여 운전자를 지원할 수 있는 중요한 기술로 많은 ADAS(Advanced Driver Assistance System) 및 자율주행 응용 프로그램에서 차량의 안전성과 정확성을 향상시키는 도구로 사 용되고 있다. LiDAR를 통해 생성된 Point Cloud Data를 처리하기 위한 다양한 방법론들이 연구되고 있으며 특히, 딥러닝 분야 에서도 활발하게 연구가 진행되고 있다. 본 연구에서는 이러한 LiDAR의 Point Cloud Data(PCD)를 처리하기 위해서 PCD의 지역적인 특징을 수학적으로 계산하고 직 접 기계에 입력할 수 있는 기법인 Topological Data Analysis(TDA)를 사용하고자 하며 이를 통해 기존의 PCD를 다루는 딥러닝 모델인 PointNet, DGCNN 등에서 발생했던 문제인 로컬 구조와 패턴에 대한 인식, 객체의 스케일과 회전 등의 변형에 대한 민 감성을 해결해보고자 한다. 이를 통해 보다 정교한 인공지능 모델을 구축하고 실제 수집된 데이터에 대해 강건하게 대응할 수 있을 것으로 기대된다.

Precise forest inventory is crucial for sustainable forest management. Recently, LiDAR technology has been widely applied to forest inventory as it provides highly accurate measurements of tree height, canopy structure, and terrain, enabling efficient data collection and analysis. This study aims to (i) determine point cloud density of ground and vegetation layers classified from Handheld Mobile Laser Scanning (HMLS), Airborne Laser Scanning (ALS), and Integrated ALS and HMLS; (ii) compare DBH and tree height derived from HMLS, ALS, and Integrated ALS and HMLS, and (iii) analyse applicability of integrating HMLS and ALS scanning methods in estimate individual tree attributes of pine forests in South Korea. As a forthcoming LiDAR application, the fusion of ALS and HMLS method could transform forest inventory methods, overcoming the challenges of traditional ground-based surveys and enabling faster and more precise forest evaluations, which contributes to forest ecosystem assessment and carbon stock estimation.

This review study discusses the utilization of Terrestrial Laser Scanner (TLS) 3D point cloud data in representing geometry for structural health monitoring (SHM) applications, particularly the so-called Scan-to-FEM method. The existing studies related to the topic are intensively presented and discussed along with the advantages and disadvantages of the modelling methods: meshing by whole and cross-section. Two case studies are conducted to show both modelling methods used for geometry representation for Scan-to-FEM method. Meshing by whole and cross-section methods result in the relative errors of respective 2.79% and 1.4% in the maximum displacement vector sum, compared to their reference models constructed by ABAQUS. It is also seen that the density and clarity of 3D point cloud data can bring well-shaped geometry for FE analysis and consequently produce good accuracy.

효과적인 산림경영계획 수립을 위해서는 수고, 흉고직경 등 수목관리정보에 대한 조사가 필요하다. 하지만 기존 의 산림조사 방법의 효율성을 향상시키기 위한 데이터 취득 기술의 융복합 및 적용에 대한 연구는 부족한 실정이다. 이에 본 연구에서는 3D 스캐너를 통해 취득되는 포인트클라우드 데이터를 활용하여 수목관리정보를 구축하고 분석하 였다. 고정형 및 이동형 3D 스캐너를 이용하여 연구대상지에 대한 데이터를 취득하였으며, 작업시간 비교를 통해 이동 형 3D 스캐너의 효율성을 제시하였다. 또한 포인트클라우드 데이터를 이용한 식생의 객체별 분류를 수행하고, 흉고직경 및 수고에 대한 정보를 구축함으로써 객체 관리가 가능한 수목관리정보를 구축하였다. 기존의 측정 방법과 비교한 정확 도 평가 결과 수고는 0.02-0.09m, 흉고직경은 0.01-0.04m의 차이를 나타내었다. 향후 추가적인 연구를 통해 객체별 식생의 위치와 수관에 대한 정보를 구축한다면 산림관리정보 구축 관련 업무 효율성 증가에 기여할 것이다.

In order to establish an effective forest management plan, it is necessary to investigate tree management information such as tree height and DBH(Diameter at breast height). However, research on convergence and application of data acquisition technology to improve the efficiency of existing forest survey methods is insufficient. Therefore, in this study, tree management information was constructed and analyzed using point cloud data acquired through a 3D scanner. Data on the study site was acquired using fixed and mobile 3D scanners, and the efficiency of the mobile 3D scanner was presented through comparison of working hours. In addition, tree management information for object management was constructed by classifying vegetation by object using point cloud data, and by constructing information on chest height diameter and height. As a result of the accuracy evaluation compared with the conventional measurement method, the difference in tree height was 0.02-0.09m and DBH was 0.01-0.04m. If information on the location of vegetation and crowns of each object is constructed through additional research in the future, the efficiency of the work related to forest management information construction can be greatly increased.

This paper presents the novel observation model, called Modified Spherical Signature Descriptor(MSSD), capable of representing 2D image generated from 3D point cloud data. The Modified Spherical Signature Descriptor has a uniform mesh grid to accumulate the occupancy evidence caused by neighbor point cloud data. According to a kind of area such as wall, road, tree, car, and so on, the evidence pattern of 2D image looks so different each other. For the parameter learning of Convolutional Neural Network(CNN) layers, these 2D images were applied as the input layer. The Convolutional Neural Network, one of the deep learning methods and familiar with the image analysis, was utilized for the urban structure classification. The case study on CNN practice was introduced in detail in this paper. The simulation results shows that the classification accuracy of CNN with 2D images of the proposed MSSD was improved more than the traditional methods' one.

LiDAR는 자율주행차에서 활용되는 핵심적인 센서로서, 3D 정보를 생성할 수 있고 다양한 환경요인에 비교적 강건하다는 점에서 활용 범위가 늘어나고 있다. 다만 LiDAR는 원리 상 강 우 시 빗방울에 의한 신호감쇠 및 산란현상으로 인해 성능이 일정 부분 저하되는 것으로 알려 져 있으며, 이에 LiDAR를 활용한 도로환경 검지 및 활용에 있어 강우에 따른 영향요인 분석이 필요하다. 본 연구에서는 강우 시 LiDAR의 성능 저하 요인이라 알려진 신호감쇠 및 산란의 영향에 대한 해석을 실측 데이터를 기반하여 진행한다. 강우량 통제가 가능한 실내 실험시설 에서 고휘도 재귀반사시트지를 활용한 시설물을 활용하여 데이터를 취득하고, 이를 신호감쇠 및 산란의 관점에서 해석하여 강우 시 LiDAR 성능 저하를 정량적으로 확인한다. 점군분포도 와 성능지표 분석결과, 비로 인한 신호감쇠와 산란의 영향으로 LiDAR의 성능은 저하된다. 세 부적으로 정량 성능지표 분석결과는 LiDAR는 비로 인한 신호감쇠 효과로 주로 Intensity가 감 소하며, 신호산란 효과로 NPC와 Intensity가 감소하고, 측정거리 오차가 커졌다.

LiDAR is a key sensor used in autonomous vehicles, and its range of applications is expanding because it can generate 3D information and is relatively robust to various environmental factors. However, it is known that LiDAR performance is degraded to some extent due to signal attenuation and scattering by raindrops during rain, and thus the need for analysis of factors affecting rainfall in road environment detection and utilization using LiDAR has been confirmed. In this study, we analyze how signal attenuation and scattering, known as factors degrading LiDAR performance during rain, cause performance degradation based on real data. We acquire data using facilities that utilize high-luminosity retroreflective sheeting in indoor chamber where quantity of rainfall can be controlled, and quantitatively confirm the degradation of LiDAR performance during rain by interpreting it from the perspective of signal attenuation and scattering. According to the point cloud distribution and performance analysis results, LiDAR performance deteriorates due to signal attenuation and scattering caused by rain. Specifically, the quantitative performance analysis shows that LiDAR experiences a decrease in intensity primarily due to signal attenuation from rain, as well as a reduction in NPC and intensity due to signal scattering effects, along with an increase in measurement distance error.

Data registration is a common operation in object digitalization. This paper proposes an algorithm for registering point cloud data with such quadratic surfaces as planes and cylinders. In the proposed algorithm, the variation of neighboring points’ normal vectors are utilized to detect data that is located at the intersection of different quadratic surfaces, which is then deleted from the point clouds. Next, region segmentation is done on remaining point clouds through the growth algorithm, and feature descriptors are created for each detected quadratic surface. Finally, rough registration of data is achieved via matching of feature descriptors, and then fine registration is achieved by optimizing rough registration results through the ICP algorithm. Experimental results show that the algorithm is insensitive to outliers, has low requirements on the original poses of data and that its registration success rate is higher than that of the original ICP algorithm.

최근 UAM(Urban Air Mobility)에 대한 관심이 도시의 교통 혼잡과 대기오염 문제 해결 방안으로 급증하고 있다. 하지만 UAM의 효율적인 운영을 위해서는 3D Point Cloud 데이터의 정확한 처리가 필요하며, 특히 지면과 객체를 분리하는 문제가 중요하다. 본 논문은 UAM 환경의 동적이고 복잡한 특성을 고려하여 지면과 객체를 효과적으로 분리하는 방법을 제안하고 검증한다. 우리의 접근 방식은 MEMS 센서로부터 얻은 자세 정보와 RANSAC을 이용한 지면 평면 추정을 결합하여, GPS 오차에 크게 영향 받지 않는 지면/객체 분리를 가능하게 한다. 시뮬레이션 결과는 이 방법이 UAM 환경에서 효과적으로 작동함을 보여주며, 도심 항공 모빌리티의 안전성과 효율성을 향상시키는 중요한 단계를 제시한다. 향후 연구는 이 알고리즘의 정확성을 높이고 다양한 UAM 환경에서 성능을 평가하며, 실제 드론 테스트를 진행할 예정이다.

Recently, interest in UAM (Urban Air Mobility) has surged as a critical solution to urban traffic congestion and air pollution issues. However, efficient UAM operation requires accurate 3D Point Cloud data processing, particularly in separating the ground and objects. This paper proposes and validates a method for effectively separating ground and objects in a UAM environment, taking into account its dynamic and complex characteristics. Our approach combines attitude information from MEMS sensors with ground plane estimation using RANSAC, allowing for ground/object separation that isless affected by GPS errors. Simulation results demonstrate that this method effectively operates in UAM settings, marking a significant step toward enhancing safety and efficiency in urban air mobility. Future research will focus on improving the accuracy of this algorithm, evaluating its performance in various UAM scenarios, and proceeding with actual drone tests.

Recently, the emergence of 3D cameras, 3D scanners and various cameras including Lidar is expected to be applied to applications such as AR, VR, and autonomous mobile vehicles that deal with 3D data. In Particular, the 3D point cloud data consisting of tens to hundreds of thousands of 3D points is rapidly increased in capacity compared with 2D data, Efficient encoding / decoding technology for smooth service within a limited bandwidth, and efficient service provision technology for differentiating the area of interest and the surrounding area are needed. In this paper, we propose a new quality parameter considering characteristics of 3D point cloud instead of quality change based on assumed video codec in MPEG V-PCC used in 3D point cloud compression, 3D Grid division method and representation for selectively transmitting 3D point clouds according to user's area of interest, and propose a new 3D Manifesto. By using the proposed technique, it is possible to generate more bitrate images, and it is confirmed that the efficiency of network, decoder, and renderer can be increased while selectively transmitting as needed.

Object recognition belongs to high-level processing that is one of the difficult and challenging tasks in computer vision. Digital photogrammetry based on the computer vision paradigm has begun to emerge in the middle of 1980s. However, the ultimate goal of digital photogrammetry - intelligent and autonomous processing of surface reconstruction - is not achieved yet. Object recognition requires a robust shape description about objects. However, most of the shape descriptors aim to apply 2D space for image data. Therefore, such descriptors have to be extended to deal with 3D data such as LiDAR(Light Detection and Ranging) data obtained from ALS(Airborne Laser Scanner) system. This paper introduces extension of chain code to 3D object space with hierarchical approach for segmenting point cloud data. The experiment demonstrates effectiveness and robustness of the proposed method for shape description and point cloud data segmentation. Geometric characteristics of various roof types are well described that will be eventually base for the object modeling. Segmentation accuracy of the simulated data was evaluated by measuring coordinates of the corners on the segmented patch boundaries. The overall RMSE(Root Mean Square Error) is equivalent to the average distance between points, i.e., GSD(Ground Sampling Distance).

Object detection and parameter estimation in point cloud data is a relevant subject to robotics, reverse engineering, computer vision, and sport mechanics. In this paper a software is presented for fully-automatic object detection and parameter estimation in unordered, incomplete and error-contaminated point cloud with a large number of data points. The software consists of three algorithmic modules each for object identification, point segmentation, and model fitting. The newly developed algorithms for orthogonal distance fitting (ODF) play a fundamental role in each of the three modules. The ODF algorithms estimate the model parameters by minimizing the square sum of the shortest distances between the model feature and the measurement points. Curvature analysis of the local quadric surfaces fitted to small patches of point cloud provides the necessary seed information for automatic model selection, point segmentation, and model fitting. The performance of the software on a variety of point cloud data will be demonstrated live.

논문은 LiDAR 점군 데이터를 사용하여 흉고 직경과 수고를 예측하는 ForestLi 시스템을 제안한다. ForestLi 시스템이 LiDAR 점군 데이터를 처리하는 과정은 다음과 같이 여러 단계로 진행된다. 다운샘플링, 이상점 제거, 지표면 분할, 지표면 정규화, 수간 추출, 개체목 분할, 흉고 직경 측정, 수고 측정. LiDAR 점군 데이터를 처리하는 상용 시스템 LiDAR360은 하측 식생과 개체목 분할 오류를 사용자가 직접 수정해야 한다. ForestLi 시스템은 하측 식생에 해당하는 LiDAR 점군 데이터를 자동으로 제거한다. 결과적으로 ForestLi 시스템이 LiDAR360보다 전체 수행시간을 줄이고, 흉고 직경과 수고 예측의 정확성을 높였다. 실험을 통해서 제안된 ForestLi가 LiDAR360 시스템보다 흉고 직경과 수고 측정의 정확성과 전체 실행시간 측면에서 우수하다는 것을 보여주었다.

In this paper, we propose a system termed ForestLi that can accurately estimate the diameter at breast height (DBH) and tree height using LiDAR point cloud data. The ForestLi system processes LiDAR point cloud data through the following steps: downsampling, outlier removal, ground segmentation, ground height normalization, stem extraction, individual tree segmentation, and DBH and tree height measurement. A commercial system, such as LiDAR360, for processing LiDAR point cloud data requires the user to directly correct errors in lower vegetation and individual tree segmentation. In contrast, the ForestLi system can automatically remove LiDAR point cloud data that correspond to lower vegetation in order to improve the accuracy of estimating DBH and tree height. This enables the ForestLi system to reduce the total processing time as well as enhance the accuracy of accuracy of measuring DBH and tree height compared to the LiDAR360 system. We performed an empirical study to confirm that the ForestLi system outperforms the LiDAR360 system in terms of the total processing time and accuracy of measuring DBH and tree height.

건설, 의료, 컴퓨터 그래픽스, 도시공간 관리 등 다양한 분야에서 3차원 공간모델이 이용되고 있다. 특히 측량 및 공간정보 분야에서는 최근 스마트시티, 정밀도로지도 구축 등과 같은 고품질의 3차원 공간정보에 대한 수요가 폭발적으로 증가하면서, 이를 보다 손쉽고, 간편하게 취득하기 위하여 MMS, UAV와 같은 관측기술이 활발히 활용되고 있다. 하지만 두 자료를 통합하여 3차원 모델링을 수행하기 위해서는, 두 관측기술 적용 시 발생하는 원시자료 취득센서, 점군 자료생성 방식 및 관측정확도 간의 차이를 효율적으로 보정할 수 있는 최적의 정합방법이 필요하다. 본 연구에서는 일반적인 3차원 모델의 자동정합에 사용되는 ICP(Iterative Closet Point) 기법을 통한 MMS와 UAV 점군 데이터 간 자동정합 성능을 판단하기 위하여, 여의도 지역을 연구대상지역으로 설정하고 UAV 영상을 취득 후 점군 자료로 변환하였다. 그 후 대상지역을 총 4개의 구역으로 구분하여 MMS 관측을 수행하였으며, UAV 점군 자료를 기반으로 각 구역에서 관측된 MMS 점군 자료와 수동정합하고 이를 ICP 기반으로 자동정합한 결과와 비교하였다. 보다 엄밀하게 ICP 기반의 자동정합 성능을 판단하기 위하여 각 구역별로 데이터 중첩률, 노이즈 레벨 등의 변수를 다르게 하여 비교를 수행하였다. 결론적으로 ICP 기반의 자동정합 시 데이터 중첩률이 높고, 노이즈 레벨이 낮을수록 더 높은 정확도로 정합될 수 있다는 것을 알 수 있었다.

3D geo-spatial model have been widely used in the field of Civil Engineering, Medical, Computer Graphics, Urban Management and many other. Especially, the demand for high quality 3D spatial information such as precise road map construction has explosively increased, MMS and UAV techniques have been actively used to acquire them more easily and conveniently in surveying and geo-spatial field. However, in order to perform 3D modeling by integrating the two data set from MMS and UAV, its so needed an proper registration method is required to efficiently correct the difference between the raw data acquisition sensor, the point cloud data generation method, and the observation accuracy occurred when the two techniques are applied. In this study, we obtained UAV point colud data in Yeouido area as the study area in order to determine the automatic registration performance between MMS and UAV point cloud data using ICP(Iterative Closet Point) method. MMS observations was then performed in the study area by dividing 4 zones according to the level of overlap ratio and observation noise with based on UAV data. After we manually registered the MMS data to the UAV data, then compared the results which automatic registered using ICP method. In conclusion, the higher the overlap ratio and the lower the noise level, can bring the more accurate results in the automatic registration using ICP method.

This study proposes an automatic registration process between Building Information Modeling (BIM) and point cloud models to enhance construction accuracy analysis. Specifically, the proposed method utilizes Z-axis and centroid alignment based on RANSAC, global registration using Fast Point Feature Histograms (FPFH), and fine registration via a 2D plane-based Iterative Closest Point (ICP) algorithm. To validate the performance of the automated process, registration results were visualized using heatmaps to analyze spatial error distributions. Furthermore, the effectiveness of the proposed technique was quantitatively evaluated by comparing the global registration stage with the complete automatic registration process using point-to-point distance error histograms and statistical metrics, including Mean, Root Mean Square Error (RMSE), and Standard Deviation. Experimental results using corridor data from a target building demonstrated that the proposed method reduced the mean error by over 80% compared to existing methods, achieving high registration precision. These findings suggest that the proposed approach improves both the efficiency and accuracy of BIM-point cloud integration, thereby streamlining construction quality control tasks while enhancing inspection precision.

As the shape and size of detectable objects diversifying recognition and segmentation algorithms have been developed to acquire accurate shape information. Although a high density of data captured by the repetition of scanning improves the accuracy of algorithms the high dense data decreases the efficiency due to its large size. This paper proposes standardization algorithms using the feature of structural members on indoor point cloud data to improve the process. First of all we determine the reduction rate of the density based on the features of the target objects then the data reduction algorithm compresses the data based on the reduction rate. Second the data arrangement algorithm rotates the data until the normal vector of data is aligned along the coordinate axis to allow the following algorithms to operate properly. Final the data arrangement algorithm separates the rotated data into their leaning axis. This allows reverse engineering of indoor point clouds to obtain the efficiency and accuracy of refinement processes.

고밀도 항공라이다(LiDAR) 3차원 점군자료는 3차원 객체 모델링, 고정밀 수치표고모형 제작, 식생 및 생태환경 분석 등 다양한 방면에서 널리 활용된다. 최근 항공라이다 장비의 고도화로 보다 고밀도의 3차원 점군자료 취득이 가능해졌으나 상대적으로 데이터의 물리적 크기가 증대됨에 따라 활용성을 확장하는데 한계가 발생하고 있다. 본 연구에서는 항공라이다 기반 고밀도 3차원 점군자료의 효율적인 활용방안을 모색하고자 3차원 점군자료의 경량화에 활용되는 대표적인 4가지 경량화 알고리즘의 주요 기능과 개별 알고리즘에 정의된 세부 파라미터 별 특성을 분석한 후 2가지 형식의 실제 실험을 통해 3차원 점군자료 경량화에 최적화된 알고리즘 선정 및 알고리즘에 정의된 세부 파라미터(Parameter)에 특성에 따라 경량화를 통해 산출된 3차원 점군자료의 품질 변화를 검증하였다. 첫 번째 실험에서 고밀도 3차원 점군자료 샘플데이터를 By density, By order, 2D grid, 3D grid 등 대표적인 4가지 경량화 알고리즘 별로 동일하게 적용하여 경량화 품질을 비교한 결과 3차원 점군자료의 물리적인 크기 감소와 원본자료의 품질유지 측면에서 3D grid 알고리즘이 3차원 점군자료의 경량화에 가장 적합한 알고리즘으로 확인되었다. 두 번째 실험에서는 3D grid 알고리즘에 정의된 2가지의 파라미터(Keep point, Operation range) 적용 실험에서 11개의 Keep point 파라미터 옵션 중 Central 옵션을 적용한 점군자료는 현지측량을 통해 취득한 모든 검사점과 비교했을 때 평균 Root Mean Squared Error (RMSE) 0.06 m 이내의 수직 위치정확도를 확보하였으며, 모든 검사점에서 오차의 편향이 거의 없는 강건한 형태의 분포를 확인하였다. 또한 Operation range 파라미터의 Grid size 옵션을 1.00 m 이내로 설정하면 원본 대비 0.03 m 이내의 상대적인 위치 정확도를 확보할 수 있었으며 원본 자료와 비교하여 객체의 경계 및 형상 유지가 가능하였다. 또한, 지붕면의 경사도와 수직 위치정확도 및 점군 자료의 감소율 간의 상관관계를 분석한 결과, 지붕면의 경사도의 크기는 수직 위치정확도와 점군자료 감소율 모두에 유의미한 영향을 미치는 종속변수임을 확인할 수 있었다. 동일한 건물에 대해 경량화가 완료된 3차원 점군자료와 원본 3차원 점군자료를 활용해 묘사를 진행하여 비교한 결과, 고층 건물일수록 건물의 경계 및 형상에 대한 판독이 용이하나 Grid size 옵션을 0.75 m 이내로 설정할 경우 대부분의 건물에 대한 묘사가 가능한 것으로 확인되었다. 묘사가 완료된 벡터데이터를 활용하여 Grid size 별로 건물 지붕면의 형상에 따른 평면 위치정확도를 검증한 결과, 건물의 경계가 단순할수록 경량화에 따른 데이터 손실에 더 민감하게 반응하여 위치 정확도 저하가 큰 것으로 확인되었다. 마지막으로, 건물 면적의 일치 여부를 확인한 결과, 건물의 면적보다는 높이가 경량화를 진행했을 때 원본 자료의 품질에 영향을 주는 요소라는 것을 확인할 수 있었으며, 건물의 높이에 따른 건물 면적의 일치율은 고층건물일 경우 98%, 중층건물은 95% 그리고 저층건물은 85%로 확인되었다.

High-density airborne LiDAR 3D point cloud data is widely used in various fields such as 3D object modeling, high-precision digital elevation model production, and vegetation and ecological environment analysis. Recently, advancements in airborne LiDAR equipment have enabled the acquisition of higher-density 3D point cloud data; however, the increased physical size of the data has posed limitations in expanding its usability. This study aims to explore efficient utilization methods for high-density airborne LiDAR-based 3D point cloud data. To achieve this, we analyzed the main functions of four representative point cloud thinning algorithms commonly used for data reduction, along with the characteristics of the specific parameters defined within each algorithm. We then conducted two experiments to determine the most optimized thinning algorithm and verify the quality changes in the simplified data based on parameter settings. In the first experiment, we applied four representative thinning algorithms-By density, by order, 2D grid, and 3D grid-to a high-density 3D point cloud data sample. The comparison of thinning performance showed that the 3D grid algorithm was the most suitable for reducing data size while preserving the quality of the original dataset. In the second experiment, we tested two parameters defined within the 3D grid algorithm: keep point and operation range. Among the 11 keep point parameter options, the 'Central' option achieved a vertical positional accuracy within an average root mean squared error (RMSE) of 0.06 meters when compared to all ground control points obtained through field surveying. It also showed a robust distribution with minimal error bias across all checkpoints. Furthermore, setting the grid size option of the operation range parameter to less than 1.00 meters resulted in a relative positional accuracy within 0.03 meters compared to the original data, maintaining object boundaries and shapes effectively. The correlation analysis between roof slope, vertical positional accuracy, and point cloud reduction rate revealed that roof slope significantly affects both vertical and reduction rates, indicating it as a dependent variable. A comparative analysis between the original and simplified 3D point cloud data for the same building showed that, while high-rise buildings allowed easier interpretation of boundaries and shapes, a grid size of 0.75 meters or less was sufficient for clear representation of most buildings. Using the vector data derived from the simplified datasets, we evaluated the horizontal positional accuracy based on roof shape and grid size. The results indicated that buildings with simpler boundaries were more sensitive to data loss from thinning, resulting in greater positional accuracy degradation. Finally, in terms of building area consistency, it was found that building height had a greater impact on the quality of the simplified data than area itself. The agreement rate of building area based on height was found to be 98% for high-rise buildings, 95% for mid-rise buildings, and 85% for low-rise buildings.