At the heart of the LiDAR revolution is its ability to transmit precise laser pulses capable of penetrating vegetation and accurately mapping terrain features on the ground. While photogrammetry often relies on aerial imagery, this approach can be hindered by the presence of dense vegetation, leading to inaccurate results due to canopy obstructions. While photogrammetry is a powerful tool for capturing terrain data, its limitations in penetrating dense vegetation to produce accurate results are significant obstacles to overcome.
When gathering precise data in densely vegetated regions, LiDAR stands out as the definitive go-to solution. Thanks to its ability to emit laser pulses capable of effectively penetrating dense foliage, LiDAR technology is uniquely equipped to uncover hidden terrain features, yielding exceptional accuracy and dependability in the process. This novel approach represents a groundbreaking advancement over traditional photogrammetric methods, which often struggle to capture the complete landscape underneath dense vegetative cover.
At its core, traditional photogrammetry relies on images captured by a camera that are subsequently used in a triangulation process to pinpoint the object’s location in space while simultaneously establishing its internal distortions and dimensions. While this method can successfully generate a detailed, three-dimensional representation of a scene, it is hindered by the significant drawback that it may only capture what the camera has observed. As a result, the digicam’s ability to perceive is limited to only capturing the uppermost layers of tree cover, which accounts for an overwhelming majority of cases, thereby defining its maximum measurable depth of subject.
Above, the yellow elements stem from a photogrammetry dataset, juxtaposed with the brown features obtained through a LiDAR scan of the same area, as illustrated in Determine 1. As evident from the data, photogrammetric factors were unable to penetrate the vegetation cover and were correctly situated above the terrain or floor. The determination of 1a as an extra instance requires a thorough analysis of the underlying data and a clear understanding of the relevant criteria.
The orthomosaic displays an extremely heavy layer of dense vegetation that obscures the underlying terrain features, with a prominent yellow stripe evident along the cross-track flight path. In the profile space, an under-exhibited photogrammetry point cloud is depicted in blue, juxtaposed with the LiDAR scan presented in purple. On this specific occasion, the sole determinants of distinct outcomes proved to be those categorized under the umbrella term “Floor”. A significant discrepancy exists at the specified site, with a notable absence of data points in the photogrammetry dataset measuring approximately 7.8 meters below ground level; the offset variability ranges between 3-4 meters above the floor.
Three exhibits demonstrate a consistent pattern in which the photogrammetry-derived point cloud appears to hover above the precise terrain, showcasing no vegetation penetration.
It’s difficult to rely on fashion models derived from photogrammetric methods with a high degree of confidence in heavily wooded regions. While this data may be applicable to open spaces and remote vegetation outcroppings where it has been collected or estimated, there is no guarantee that it accurately reflects the underlying terrain. Attempting to survey a terrain in the manner described will produce irrelevant subsets of data similar to those generated by Digital Terrain Models (DTMs) and contour maps, yielding no meaningful insights.
In conclusive terms, the application of LiDAR technology far surpasses the limitations of its image-only photogrammetry predecessor. While these low-cost methods can facilitate data collection for DTM or contour mapping, their outputs are often inaccurate, providing a misleading representation of the terrain that can lead to significant disparities in downstream calculations by the user.
Uncover extra from sUAS Information
Sign up to receive our latest posts directly to your inbox.
