Pros and Cons of Structure for Motion Embarked on a Vehicle to Survey Slopes along Transportation Lines Using 3D Georeferenced and Coloured Point Clouds

Abstract

We discuss the different challenges, pros and cons of the fairly new Structure for Motion (SfM) embarked on a vehicle (SfM-EV) technique for slope surveys along transportation network tracks using action cameras embarked on standard moving vehicles. This low-cost technique generates georeferenced and coloured 3D point clouds without using any ground control points. Four action cameras, two of which had an integrated GNSS chip, were used to collect a series of pictures of tracksides at a rate of two images per second each. The SfM-EV results were compared with the results of seven other 3D survey techniques to evaluate the precision and accuracy of this technique, demonstrating the ability of this simple setting to generate 3D scenes. Different platforms for the cameras were tested, such as a bike, car, train, funicular, helicopter and so on. The SfM-EV technique was also tested on several study sites to highlight its strengths and weaknesses and obtain data, such as the density of points, equations of errors, overlap ratios and image resolution. The precision of the SfM-EV results was sufficient for detecting topographical changes close to the track for a volume of ~1 dm3 and the absolute positioning obtained with a low sky obstruction was approximately 5 m. The precision of SfM-EV was of a similar order to the other techniques, with an order of magnitude of a few centimetres. This approach possesses a low price-result quality ratio and is very simple to use. The possibility of using any type of vehicle for surveying is an advantage, especially for transportation track embankments.

3D Point-Cloud Processing Using Panoramic Images for Object Detection

“…Additionally, our method offers potential for broader applications beyond traditional 3D point-cloud processing domains. The integration of panoramic images [16], [17] with laser [18], [19]…”

IMG2nDSM: Height Estimation from Single Airborne RGB Images with Deep Learning

“…Structure from motion is a technique for estimating 3D structures from 2D image sequences. The main disadvantages of SfM include the possible deformation of the modeled topography, its over-smoothing effect, the necessity for optimal conditions during data acquisition and the requirement of a ground control point [15]. Like SfM, DSM estimation by stereo image pairs requires difficult and sophisticated acquisition techniques, precise image pairing and relies on triangulations from pairs of consecutive views.…”

Self-supervised monocular depth estimation from oblique UAV videos

“…A set of oblique view images with sufficient overlap, acquired from UAVs can be used for reconstructing a dense 3D model through photogrammetric image techniques like Structure-from-Motion (SfM). The general steps involve feature detection and matching by finding correspondences between images, sparse reconstruction, bundle adjustment followed by 3D dense point clouds generation (Vallet et al, 2012;Voumard et al, 2018;Nex et al, 2015). However, these imaging techniques suffer from issues like requirement of multiple images of the same scene, time consumption, low-quality generation in regions where the camera has a narrow view and for images captured along camera optical axis.…”