10:45 AM - 12:15 PM
[STT40-P01] Development of portable device for topographic measurement and its 3d-modeling
Keywords:topographic measurement, 3D model, point cloud data
It is valuable to measure surface rupture along an earthquake fault soon after a large earthquake and topographies in other field surveys. Topographic survey using air planes or drones are being developed and they contribute to various researches. Small portable devices such as tablets and smart phones can construct 3D-models of small areas. In these technologies, distance and direction measurement and graphic measurement are used. We are developing a device using low-coast LiDAR and cameras to measure topographies and to make its 3D model. Point cloud data is obtained with the LiDAR and the obtained image data is used for texture mapping of the 3D model
We use a RPLIDAR A1M8, which can measure distance up to 12 m and has ability of getting 8000 points/s. The RPLIDAR A1M8 measures distance on a plane. By rotating the LiDAR with a servo motor, we can get point cloud data in half space. Image of half space can be got with four wide-view cameras. Two Raspberry Pis are used to control and data acquisition (one for the LiDAR, and one for cameras). Two large-capacity mobile batteries supply power for continuous operation more than eight hours. This device is aimed at construction of a 3d-model for large area with easiness such as a tablet.
The point cloud obtained from LiDAR can be used to construct the framework of the model. Indoor tests showed difficulties in getting distance data for the planes of large incident angles. It is necessary to relate each pixel of image data to the coordinate to use image data for 3d modeling. We measured the direction of the image center and camera rotation angle by getting images of graph paper. Image correction was done only for the angle from the camera center direction. As preliminary tests, we got data with static condition and constructed a 3D-model from the obtained data. OpenCV library was used to manipulate image data, and OpenGL library was used to construct the 3D-model.
We use a RPLIDAR A1M8, which can measure distance up to 12 m and has ability of getting 8000 points/s. The RPLIDAR A1M8 measures distance on a plane. By rotating the LiDAR with a servo motor, we can get point cloud data in half space. Image of half space can be got with four wide-view cameras. Two Raspberry Pis are used to control and data acquisition (one for the LiDAR, and one for cameras). Two large-capacity mobile batteries supply power for continuous operation more than eight hours. This device is aimed at construction of a 3d-model for large area with easiness such as a tablet.
The point cloud obtained from LiDAR can be used to construct the framework of the model. Indoor tests showed difficulties in getting distance data for the planes of large incident angles. It is necessary to relate each pixel of image data to the coordinate to use image data for 3d modeling. We measured the direction of the image center and camera rotation angle by getting images of graph paper. Image correction was done only for the angle from the camera center direction. As preliminary tests, we got data with static condition and constructed a 3D-model from the obtained data. OpenCV library was used to manipulate image data, and OpenGL library was used to construct the 3D-model.