The rapid growth of BIM in the construction engineering industry, mandatory use in government procurements and utilization for improving productivity, the importance of 3D product modeling in the civil engineering field is becoming increasingly prominent. This study proposes a new management method for the government to use a 3D product model based solution to support EVM, objectively evaluates the government's procurement unit and multiple construction projects. Firstly, a data structure was defined regarding the relationship of the work breakdown structure (WBS) in the government contract units by using a 3D product model. Then, an EVM system was developed to calculate quantity, cost and working day. Finally, a 3D steel bridge product model was used to verify the effectiveness of the system by using a newly defined data structure.

View full abstract

 Large-scale earthquakes are often accompanied by simultaneous multiple fires. In addition to the collapse of buildings due to the earthquake, inundation by the tsunami, etc., it is necessary for local governments and residents to understand the risks of fires properly. It helps not only the initial fire fighting and evacuation by residents in case of fires, but also the town planning against disasters and actions by fire stations.
 Although the spreading fire analysis can be used for the risk assessment in earthquake fires, there is a limit in analyzing a large area collectively due to the memory of computers and calculation time.
 In this study, the method dividing the target area into blocks and integrating the results of each block is proposed in the spreading analysis of using the electronic residential map. It is applied to the risk assessment of the earthquake fire.

View full abstract

 The importance of the education on the disaster prevention has been strongly recognized since the Great East Japan Earthquake March 2011, and the education has begun on the correct knowledge of the earthquake and tsunamis and understandings of the behavior in the events of disasters. Various texts and contents on the education on the disaster prevention gave been developed, but sufficient consideration has not been made on the motivation and the effect. In this research, RPG software is created that enables to learn the manual for major earthquake produced by National Institute of Technology, Wakayama College. Then, comparing the percentage of correct answers of the tests conducted twice on the game learning group and the text learning group, the validity of the game learning is discussed.

View full abstract

 In the public space including railway stations, a signage is provided for the purpose of facilitating the movement of people. In the study on signages, walking experiments in reality space are mainstream, but problems remain in those methods. In this study, investigating whether the visibility of hanging signage can be evaluated by acquiring railroad station space as 3 dimension point cloud data with MMS and performing simulation. In evaluating the visibility, the influence due to the degree of congestion of pedestrians and the influence due to the difference in viewpoint height are taken into consideration. As a result, it is possible to evaluate the visibility of the signage in the virtual space on the 3D point cloud data by comparing the visibility of a signage with the viewpoint of the healthy person and the viewpoint of the wheelchair user.

View full abstract

 Japan has developed QZSS, and operations will be started with four satellites from fiscal 2018. In this study, we verified the improvement effect of the positioning accuracy by using GPS and three "MICHIBIKI" on quasi-zenith satellite orbit (QZO). We conducted a 24 hours observation experiment in different surrounding environment and compared them. As a result, it was confirmed that the positioning accuracy was most improved especially in vertical direction when using GPS and three QZSS. When using three QZSS, one or more satellites always exists at an elevation angle over 70°. And, QZSS are especially effective on places where influence of the obstructions is large, because QZSS near the zenith are less susceptible to shields.

View full abstract

 Monitoring of urban structure is required against the declining birth rate and population aging. The distributions of housing vacancy are expected to be one of the indicators of monitoring low-used and unused spaces around urban areas. However, there are many discussions on the time and labor problems in the field survey and the method of identifying vacant houses. In this study, the field survey of vacant houses was conducted in 2016 and 2017. We also applied the ordered logit model with hierarchical Bayesian method by using the two-time field survey results, water hydrant data, fixed assets data, local demographics, and building data. The accuracy of the detection of vacant houses was discussed through the comparison of the other findings of vacancy.

View full abstract

 In urban areas, it is expected that the number of businesses such as reconstruction, renewal, and the construction of new structures will be increased in areas where there is a large volume of remarkably deteriorated infrastructure equipment and underground buried objects. However existing structures are mainly managed by two-dimensional drawings, it is difficult to determine the position and height of underground objects during construction works.
 The purpose of this research was to devise a simple 3D model generation method of underground buried objects and establish a method to correct existing 2D drawings. In this research, the authors have tried to generate a 3D model using smartphones and digital cameras, which are recently developing and spreading widely, to examine the possibility of its application to the condition of underground buried objects for the construction planning and the temporary works planning. As a result, it was confirmed that the generated 3D model has positional accuracy that can be used for correction of 2D drawings.

View full abstract

 To grasp railway clearance is important for the safe operation of trains. In view of the difficulties to inspect railway clearance rapidly and precisely from actual positions of rails, this paper proposes a methodology for extracting the track center point of rail clearance gauge and visualizing the clearance gauge by using with Mobile LiDAR Systems directly above the rail. First, two MMSs are settled on directly above rails for data acquisition of the center positions of the railhead. Since the original reference point of the clearance gauge is the middle point of the level line on top of the two rail heads, it is calculated and evaluated using point cloud in every scanline of the laser scanner of MMS. Second, the prototype program has been developed to visualize three dimensional models of the expanded clearance gauge which is consecutively stepped down from curve to straight line along the curve track using center points. Finally, this method was applied to the test rail line where an obstacle was installed 5m up above the rail and it was identified within 0.02m precision inside of to the clearance gauge. The applicability of this method has been verified to both curve and straight tracks.

View full abstract

 A result of a visibility analysis on GIS can indicate landscape factors objectively. However, subjective views of individuals are also required in landscape design processes. To analyze a landscape subjectively, eye tracking systems have been drawing research interest because of their potentials in quantification. The development of hardware technology and software technology of eye tracking system has offered a new perspective and study platform for the study on landscape visual perception. Under these circumstances, the authors are trying to utilize the eye tracking system to analyze interests of participants, and trying to compare with the result of the visibility analysis on GIS. A screen-based eye-tracking system was selected because of its short calibration time. The railway overpass project in Neyagawa city has been selected as a case study.
 In this study, the authors proposed the method that can generate wide-area model efficiently by utilizing GIS functions and various spatial data. Then, two types of landscape analyses were carried out on GIS. Precisely, the visibility analysis was performed by using representative points of the railway viaduct and DSM which was converted from the urban model. As the result, the influence of the viaduct on the landscape and the future landscape from the railway viaduct were clarified quantitatively by visibility analysis. In addition, the difference between the result of visibility analysis on GIS and the actual gaze trend were confirmed by utilizing the eye tracking system. From the results of this study, it can be said that there is not much relation between the visibility (calculated on GIS) and the actual gaze trend (analyzed by eye tracking system). Buildings located near the viewpoint are not gazed at well, even if their visual size are large. In addition, Objects obstructing the skyline are paid attention to.

View full abstract

 In Japan, to support the realization of a self-driving system and a safe driving system, development of a highly accurate dynamic three-dimensional map is being promoted. Dynamic mapping consists of 34 kin ds of real features and virtual features. These features differ in specifications, such as the usage and position acquisition, which are present in fundamental geospatial data of road.
 Therefore, when mutual conversion between them becomes possible, freshness and exhaustiveness can be guaranteed, which can be expected to contribute greatly to the mutual agreement of both.
 In this study, the affinity was analyzed of the features and attributes of dynamic map and fundamental geospatial data of road. Then, a conversion method was devised for 12 features with high affinity, and the usefulness was clarified through a case study.

View full abstract

 In concrete placement, it is extremely important to transfer the empirical knowledge of skilled workers to inexperienced workers, because there are many manual-based operations, such as the operation of construction instruments. In the Japanese construction industry, succession of skills is an urgent issue because it is expected that many skilled workers will retire in succession due to the aging progresses. Specifically, there is the possibility that empirical knowledge related to the behavior of skilled workers can be handed over to the next generation by measuring and accumulating it and converting it to explicit knowledge. However, it is not clear whether this measurement is possible or not.
 The purpose of this study is verify the feasibility of the measurement of empirical knowledge related to the behavior of skilled workers by using various measurement techniques. In this paper, we report the measurement method of the insertion position and timing of vibrator related to concrete placement in relation to skill succession and the measurement of differences according to each individual.

View full abstract

 In recent years, the development of automatic inspection robot for the cables of cable-stayed bridges has become active. Therefore, the authors have developed an automatic inspection robot named CHiQWA for the cables of cable-stayed bridges. This robot has four features. The first is able to reach the top of the cable using the UAV. The second is able to perform descent of the inspection robot by automatic control. The third is able to capture takes continuously scanned images of cable surface with USB cameras, miniPC, and lighting devices. The fourth is able to generates a panoramic image of cable surface. Actually, the existing robot is not able to create an accurate panoramic image due to circumference of a cable. Therefore, in this study, the author developed the image corecting method for expand circumference by projective transformation. Furthermore, by applying this method, we found the possibility of circumferential expansion by projective transformation.

View full abstract

 This article examined efficient method for falling rock investigation using tablet type GIS and microtopography highlight map. In the experiment field, it was possible to extract a steep cliff with a height difference of about 1.5m or more and an inclination angle of 60° or more from microtopography highlight map, but it was also confirmed the necessity of field investigation including risk judgment. In the field investigation, we verified the relationship between the microtopography highlight map and the self position indicated by the tablet terminal at the slope of the mountain, and it was effective for approaches to falling rock source. Moreover accumulation of investigation data using tablet type GIS not only contributes to the efficiency improvement of disaster prevention record update but also it is effective for future maintenance because it also has position repeatability of falling rock source.

View full abstract

 There are many natural disasters caused by earthquakes and heavy rains in Japan, also human injuries and environmental damages have been a big problem. In the present situation, in order to grasp the distribution of landslide, the worker create landslide maps by slope failure and debris flow from aerial photographs took after landslide disaster. However this work requires a detailed visual inspection for a long time which places a heavy burden on workers. Therefore, in order to improve efficiency of creating landslide maps, we took advantage of the method of automatic detection of landslide from aerial photographs by deep learning. In addition, the images to be processed in this research are aerial photographs were targeted for grayscale images which not be archived.

View full abstract

 The number of vehicles equipped with ETC 2.0 navigation exceeds 4 million units, and the data acquisition amount of ETC 2.0 probe data has dramatically increased. The authors are working on research and development of a method for understanding the total amount of traffic flow by ETC 2.0 probe data. In order to understand the total amount of traffic flow, it is necessary to detect the origin-destination point of the trip according to the road traffic situation, but ETC 2.0 probe data includes incorrect origin-destination points.
 In response to this problem, we propose a method to automatically identify and correct incorrect origin-destination points included in ETC 2.0 data. Specifically, links which are considered to be unlikely to be origin-destination points in the road network are defined as a “no origin-destination link”, all origin-destination points existing on no origin-destination links are regarded as incorrect origin-destination points and corrected. We tried correction of incorrect origin-destination points by the proposed method and confirmed its usefulness.

View full abstract

 Recently, due to the inundation damages in urban areas caused by frequent torrential rainfalls in short-time, the risk of flooding in the underground malls in urban areas is also increasing. Analysis of groundwater flooding damage predicted from the rainfall situation is progressing. However, the method of communicating predicted risk of inundation to the general public in the target underground space is unresolved. This paper proposes a method to display the predicted depth of flooding with graphical 3D CG (Computer Graphics). Distribution of 3D data and rendering of CG increase the load on calculation and communication at the terminal, so the server generates an all-around panoramic image matched with the user's position and distributes it as HTML (Hyper-Text Markup Language) for normal web access in the proposed method. This method makes it easier to understand the predicted inundation situation, and not only it can be expressed from the first-person perspective but also it is possible to check the situation interactively by looking around the surroundings. This paper shows the effectivity of our method by experimental implementation at an actual railway station facility in underground and inundation simulation results based on assumed heavy rainfalls of past data in Osaka city.

View full abstract

 In recent years, many of the infrastructures have been exceeding their useful life, and intensive maintenance is required. In order to incorporate a UAV as an effective alternative to direct visual inspection, it has to visit all the inspecting points and take photos of them in proper resolution. This paper proposes a method for making a flight path that allows taking proper photos of target structures cost-effectively with mathematical programming technique. The proposed method first identifies the optimal viewpoints which cover all the inspecting points and satisfy the overlap ratio of photos. Then, the shortest cyclic path can be obtained to visit all the optimal points as the optimal flight plan. The numerical experiments showed the quantitative performance of the proposed method for planning and visualization of the optimized flight path.

View full abstract

 At the construction fields, the aging of skill engineers and the lack of successors are growing into a serious problem. And the increase in the number of labor accidents due to crushing labor is becoming a growing concern. Therefore, the thorough safety control of the workers is the urgent action assignment. To decrease the number of labor accidents, it is essential to detect the near miss incidents which can be directly linked to accidents, also to understand the physical fatigue condition of individuals at all time. Looking at existing cases, there are researches aimed for the decrease of the road traffic accidents. One is to detect near miss incidents of the drivers using the vehicle behavioral data such as an acceleration rate. Another research is to estimate the physical fatigue of the driver using the biological information such as the heart rate. However, these cases are intended for persons in sitting position, which cannot be applied to the workers at the construction fields.
 Therefore, in this research, we validated the various possibilities concerning the near miss incidents and the physical fatigue using smart watch, which can easily measure the biological information at a reasonable price, under the environmental conditions considering the construction fields.

View full abstract

 In road traffic analysis using probe data capable of identifying future driving lanes, lane network data prepared for the purpose of driving support is necessary. However, there is no prospect of preparing and updating lane network data in the areas required for road traffic analysis. Therefore, if fundamental geospatial data of road being developed for road management purposes can also be used for road traffic analysis, in addition to expanding the use of maps and manifesting the effect of improvement, the data of the planning and management system of the road can be used. This is also expected to contribute to improving interoperability.
 In this study, the authors evaluated the satisfaction level of a large-scale road map suitable for road traffic analysis for fundamental geospatial data of road, and examined countermeasures against the requirements which were not reached. Furthermore, based on these countermeasures, existing road-based map information was expanded and the usefulness was verified by a case study using high-accuracy probe data and large-scale road map.

View full abstract

 This paper has proposed a strategy for supporting real-time simultaneous analysis of hyperspectral cube moving image (termed HSC moving image) and spectral reflectance characteristics. Based on a visual illusion-based image enhancement system (VIS) newly developed in this research, an animation-like Feature Composite moving image (FC image) inducing visual illusion with the emboss processing can be produced. The power spectra for the fluctuation of digital number values (i.e., spatial frequency components) in each pixel of FC image is also delineated on "Visibility Evaluation image (VE image)." Through the experiments on concrete surface texture analysis, the results indicate as follows: i) The real-time simultaneous analysis of HSC moving image and spectral reflectance characteristics can be implemented using VIS with user interface superior in operability and; ii) furthermore, for supporting selection of training data sets to produce an image feature classification, the HSC moving image itself can be provided as one of the newly products.

View full abstract

 From the viewpoint of effective utilization of infrastructure, in order to improve user satisfaction level, measures for improving comfort at the time of road traveling have been emphasized in recent years. It is thought that road condition and scenery are related to road comfort.
 In this study, a comfort map was constructed as a measure to improve comfort. The comfort map was constructed based on the sensitivity evaluation experiment of a specific section of Hakone turnpike. For the first landscape, the category score analyzed by quantification theory is used. For the second road surface condition, the value of the International Roughness Index is used. Based on these two values, they are plotted on the map as a comfort map. This comfort map has both a road manager's perspective and a road user's perspective. Road managers use this comfort map for maintenance. Road users are used to motivate visits.

View full abstract

 In this paper we propose the LOD (Level of Development) of the three-dimensional model concerning the sluice gate based on the business process of the sluice gate, and confirm the validity of the proposal by case study of the two sluice gates. We suppose the expected effect by utilizing the three-dimensional model in each business process of the sluice gate, and confirm a part of the expected effect of the utilization by trial in the river sluice gate that is in the construction stage.

View full abstract

 Once the earthquake occurs, the staff of the disaster management bureou of government agencies effort to clarify damage to infrastructure. CCTV (Closed Circuit Television) cameras have developed and have been utilized to survey infrastructure on roads and rivers. We have been studying an automatic system to select CCTV cameras in municipal area, where seismic intensity exceeds a preset level and to render panoramic images by stitching camera images obtained during panning CCTV cameras.
 The earthquake can occur under bad lighting conditions such as night. During panning CCTV cameras at night, the bands of light (tails of residual image) are observed in images resulting from electron multiplier tube to obtain the amount of light. The bands of light are noises to render panoramic images.
 In this paper, therefore, we propose a method to panning and stopping little by little. The effect of electron multiplier can be different from the type of CCTV cameras. In addition, the conditons around the CCTV cameras can affect the amount of noises (the bands of light). Therefore, we applied our method to CCTV cameras located on different municipal area.
 Panoramic images are necessary soon after the earthquake to support the early stage. We validated our method by measuring the stopping time to remove the bands of light. As a result of trial for 16 domestic CCTV cameras, we clarified that almost 3 seconds are adequate to remove the bands.

View full abstract

 Indoor navigation has problems such as lack of unified positioning method and undeveloped three-dimensional map, and it is not practically used. In this study, point cloud data was acquired in a building using indoor type MMS. As preliminary examination of 3D map maintenance from point cloud data, shape measurement and creation of 3D model were performed. As a result, three-dimensional model created from point cloud data based on geocentric geodetic datum showed possibility to be used as a map for indoor navigation.

View full abstract

 The Global Navigation Satellite System (GNSS) is expected to be utilized in construction fields such as civil engineering construction. It is possible to use in urban areas in the future, but the data obtained by GNSS tends to decrease accuracy in places where shields such as high-rise buildings exist etc. In this research, in the shielding environment such as this In order to improve the accuracy of high precision multi GNSS data, we aim to verify the satellite selection method based on visibility evaluation using 3 dimensional urban model. As a result of analyzing the acquired data in the shielding environment from the ratio of Fix solution, improvement of accuracy was confirmed according to shielding conditions.

View full abstract

 This study aims to verify the change of the positioning accuracy by comparing the positioning accuracy focusing on the difference of the number of QZSS and the elevation angle of QZSS during the positioning with GPS and QZSS. Comparing the accuracy with the increase of number of QZSS, the accuracy improved. When the satellites were located at high elevation angle, the accuracy tended to be high, however the number of satellites and DOP were also greatly related to the accuracy.

View full abstract

 A point cloud based visualization system of sewer pipe networks was developed based on point cloud data in order to reduce their maintenance and operation costs. Two types of point cloud data were generated. The first type is a point cloud date set obtained via images taken mainly by a UAV mounted camera flying over a college campus. The second type is a data set generated from 3D CAD models of sewer pipe networks found in the campus. These two types of data sets were merged into one hybrid data set by which the sewer networks can thus be visualized in relation to buildings, roads, and other objects above the ground. Due to the huge file size of the point cloud data set, instead of using it directly, alternative video clips were generated by displaying the point cloud on a PC using a point cloud viewer and capturing displayed scenes on various view points in video. By referencing these video clips stored in a data base accessed by a smart phone via the Internet, one can understand how sewer pipes are connected to each other or which manholes are linked to which manholes via pipes. The data base also stores the results of sewer integrity studies such as inspection video clips and damage lists. The developed system can present automatically proper section of sewer pipe networks based on current locations calculated from GPS signals received by a smart phone. Hence the corresponding video clip of point cloud and inspection results are shown on the smart phone.

View full abstract

 In this study, we have developed a simulation system of the flood disasters such as sediment flood, tsunami, storm surge and so on, by using a game engine and a virtual reality device. To calculate the flood flow, we calculate the free surface flow based on the three-dimensional VOF method using OpenFOAM solver. In addition, we adopted the debris to the simulation system. The calculation of the motion of debris such as driftwood and rock are carried out using the physical engine of the game engine. By reflecting the motion of debris to the virtual reality space, a high-quality simulation of flood disasters has become possible.

View full abstract

 Volcanic activity of the Kilauea volcano on Hawaii Island has been activated since May 2018, and a large eruption continues to occur. These volcanic eruptions have caused the lava flow to drain, volcanic ash and volcanic gas-containing plume, which has caused a lot of damage to buildings, roads and the surrounding environment. In this study, we investigated the spread of lava flow due to eruption, the change of shape of crater and the situation of crustal deformation due to earthquake, using Sentinel-1 observation information of C band satellite SAR. The shape of the crater, changes in the surface of the ground due to lava flow and crustal deformation were effectively evaluated. The applicability to overview survey in disaster quickly response and continuous monitoring was considered from the survey results using these Sentinel-1 satellite images with all-weather characteristics that can be obtained in near real-time after satellite observation. As a result, the usefulness of the Sentinel-1 satellite was confirmed for the investigation of lava flow distribution and the diastrophismt due to volcanic eruption.

View full abstract

 Recent years, the number of potential wheelchair users is acceleratingly growing in Japan, as facing the issue of rapid aging society. Barrier-free maps, for example, are prepared in many municipalities, but they miss the details of physical barriers, such as bumps and width clearances on the path, needed for the independent mobility of the wheelchair users. This paper addresses a method to find out physical barriers by using a depth camera, so the facility managers easily check and improve the environment by themselves. The proposed system robustly estimates the floor surface from the point cloud captured by a depth camera and RANSAC algorithm and highlights the existence of interference between the environment and the volume of an actual wheelchair would occupy to pass through. This paper demonstrates the achieved accuracy of the barrier checks in an actual public space.

View full abstract

 Ministry of Land, Infrastructure, Transport and Tourism in Japan is promoting the intelligent construction called “i-Construction” for productivity improvement, quality assurance, and cost reduction. Three-dimensional landform data are generated by laser profiler, mobile mapping system, terrestrial laser scanner, and camera on unmanned aerial vehicle. However, there are no standards for representing and drawing three-dimensional landform data.
 In this study, the representation and modeling methods of three-dimensional landform data based on point cloud data were proposed. This study evaluated the density of point cloud data and its display for representing point cloud data based on various measurement hardwares. The density was evaluated for asbuilt management of road earthmoving. It should be 100 points/m² for as-built management. And, points of attention were considered for combining and using point cloud data.

View full abstract

 The Ministry of Land, Infrastructure, Transport and Tourism locates three-dimensional geospatial information as the most important foundation of the nation and promotes development of technology for efficiently and rapidly preparing data. However, in indoors or underground spaces, the accurate location can't be determined with GNSS. In recent years, several mobile sensing units have been developed and used. In exsinting products and researches, SLAM which calculates the relative position and attitude change amount of the measurer from time series data of images and point cloud data and generates three-dimensional data of the space has been proposed. However, in SLAM, self-position can't be estimated under the shape features such as breaklines and the environment with little change in brightness value, and It's difficult to generate three-dimensional data. In this research, a mobile sensing unit was developed for measuring three-dimensional data in indoors. A method to improve the accuracy of generating point cloud data was proposed. It estimates the planar region from the three-dimensional data measured using the unit and generats a breakline at the intersection between the planes.

View full abstract

 In recent years, the use of tablet terminals has spread in fieldwork. In this re-search, we focus on cameras and handwriting input functions standardized on tablet terminals. In this research, we propose a system that associates handwrit-ten information with AR (Augmented Reality) function and aims to enable smooth transmission and sharing of information. In the proposed system, a user is supposed to write attention points and/or draw figures directly on the screen-shot of the site. The handwritten information pops up automatically overlaid on the display when the same scene is view with the camera. In this system, the us-er saves handwritten contents on site and extracts and registers natural feature points from the camera image. Markerless AR can be realized by recognizing registered natural feature points and superimposing and displaying handwritten information. In this paper, assuming the scenario of inspection work for con-crete structures, we implement two types of AR indications, the indication of place and always oriented towards the user, and report on experimental display performance and effect.

View full abstract

 It is expected that point cloud data generated by UAV photogrammetry will be utilized in various management scenes such as construction management and disaster. On the other hand, in UAV photogrammetry, there are problems that require enormous time to generate point cloud data and that can not be measured except in bright daytime hours. For these reasons, laser scanner equipped UAV capable of measuring point cloud data at night and generating them in less time than UAV photogrammetry have been developed. However, since the accuracy of generation of point cloud data with the laser scanner equipped UAV varies depending on the conditions of the construction fields and the measurement method, it is necessary to realize the accuracy under various conditions. In this research, the measurement error of the UAV mounted laser scanner by the environment and measurement method are investigated and clarified at the viewpoint of the influence on the generation accuracy of each error factor in the sites of construction management. And the accuracy of generation of point cloud data with measurement method that takes measurement error into account is clarified.

View full abstract

 Conventionally, measurements of width, length, and height have been performed using tape measure and level in management of work done. Since 2016, the Ministry of Land, Infrastructure and Transport is working on i-Construction and aims to improve productivity. Among them, in April 2016 formulation management procedure (draft) (using pavement work) using a terrestrial laser scanner was announced, and it is expected that usage will increase in the future. Although the result could be measured with the accuracy according to the performance of the model, there were cases in which it was not possible to secure the assumed precision due to the malfunction of the equipment and the observation condition.

View full abstract

 In recent years, UAV (Unmanned Aerial Vehicle) attracts attention as a method to easily photograph aerial images with low ground resolution. Three-dimensional modeling software using SfM (Structure from Motion) and MVS (Multi-View Stereo) is generally used for three-dimensional measurement using this UAV aerial images. However, it is pointed out that these trends are different from each other. Therefore, in this research, we extracted features from two experiments, Pix4D and PhotoScan, which are currently in wide use, and compared the differences between them. As a result, we found that the lens distortion correction is different in each software, PhotoScan is beautiful in appearance but weak in displacement extraction, Pix4D is swaying in Z value but strong in displacement extraction etc.

View full abstract