Forest Stand Segmentation Using Airborne Lidar Data and Very High Resolution Multispectral Imagery

Dechesne, Clément; Mallet, Clément; Bris, Arnaud Le; Gouet, Valérie; Hervieu, Alexandre

doi:10.5194/isprsarchives-xli-b3-207-2016

Cited by 8 publications

(10 citation statements)

References 17 publications

(19 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The detection of individual trees, not of their species, was the target of other previous works [32][33][34]. Given the complex environment of trees in a forest, most authors mainly use hyperspectral cameras and LiDAR or laser-scanner sensors to build photogrammetric point clouds and count the dominant tree crowns.…”

Section: Introductionmentioning

confidence: 99%

On-the-Fly Olive Tree Counting Using a UAS and Cloud Services

et al. 2019

View full text Add to dashboard Cite

Unmanned aerial systems (UAS) are becoming a common tool for aerial sensing applications. Nevertheless, sensed data need further processing before becoming useful information. This processing requires large computing power and time before delivery. In this paper, we present a parallel architecture that includes an unmanned aerial vehicle (UAV), a small embedded computer on board, a communication link to the Internet, and a cloud service with the aim to provide useful real-time information directly to the end-users. The potential of parallelism as a solution in remote sensing has not been addressed for a distributed architecture that includes the UAV processors. The architecture is demonstrated for a specific problem: the counting of olive trees in a crop field where the trees are regularly spaced from each other. During the flight, the embedded computer is able to process individual images on board the UAV and provide the total count. The tree counting algorithm obtains an F 1 score of 99 . 09 % for a sequence of ten images with 332 olive trees. The detected trees are geolocated and can be visualized on the Internet seconds after the take-off of the flight, with no further processing required. This is a use case to demonstrate near real-time results obtained from UAS usage. Other more complex UAS applications, such as tree inventories, search and rescue, fire detection, or stock breeding, can potentially benefit from this architecture and obtain faster outcomes, accessible while the UAV is still on flight.

show abstract

Section: Introductionmentioning

confidence: 99%

On-the-Fly Olive Tree Counting Using a UAS and Cloud Services

et al. 2019

View full text Add to dashboard Cite

show abstract

“…For delineating patches with homogeneous properties with respect to ecological process functions, such as similar canopy interception probabilities and water balance conditions, the vertical stand structure becomes explicitly relevant [4,26], and its incorporation into the delineation process is a requirement. This represents a novelty compared with existing approaches since, traditionally, forest stand delineation is performed through visual interpretation of high-resolution aerial photographs by human operators considering canopy height, canopy structure, and tree species [18,27]. Manual image analysis, however, is a time-consuming task, limited by artifacts in data as, e.g., shadow effects, prone to errors and partly subjective [18,28,29].…”

Section: Introductionmentioning

confidence: 99%

“…Stand delineation in computer vision and image analysis amounts to a segmentation problem [27], the classic interpretation of which is the division of an image into homogeneous and spatially contiguous regions [33]. An early forest stand delineation approach using ALS data by Diedershagen et al [18] was based on the canopy heights derived from terrain-normalized ALS point clouds.…”

Section: Introductionmentioning

confidence: 99%

“…As is apparent, these studies mainly focused on tree heights, density, maturity, and dominant tree species or types [24,27]. However, forest management concepts that favor a more diverse forest structure or succession processes on a landscape scale may lead to diversification in tree age and height, a more distinct layering, and the presence of multiple tree species within forest stands [19,36].…”

Section: Introductionmentioning

confidence: 99%

“…In order to confirm the robustness of the approach, the delineation was performed for test sites that cover a range of forest characteristics, including forest type, management, and species composition, and have different terrain conditions regarding topography and terrain elevation. The initial stand delineations in feature space are often spatially non-contiguous and require post-processing steps (e.g., [27,40]) in order to meet the requirement of spatial contiguity of segments in terms of image analysis [33] or stands in forestry [23,34], respectively. Our framework delivers the appropriate input for such further processing.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Adaptive Framework for the Delineation of Homogeneous Forest Areas Based on LiDAR Points

et al. 2019

View full text Add to dashboard Cite

We propose a flexible framework for automated forest patch delineations that exploits a set of canopy structure features computed from airborne laser scanning (ALS) point clouds. The approach is based on an iterative subdivision of the point cloud using k-means clustering followed by an iterative merging step to tackle oversegmentation. The framework can be adapted for different applications by selecting relevant input features that best measure the intended homogeneity. In our study, the performance of the segmentation framework was tested for the delineation of forest patches with a homogeneous canopy height structure on the one hand and with similar water cycle conditions on the other. For the latter delineation, canopy components that impact interception and evapotranspiration were used, and the delineation was mainly driven by leaf area, tree functional type, and foliage density. The framework was further tested on two scenes covering a variety of forest conditions and topographies. We demonstrate that the delineated patches capture well the spatial distributions of relevant canopy features that are used for defining the homogeneity. The consistencies range from R 2 = 0 . 84 to R 2 = 0 . 86 and from R 2 = 0 . 80 to R 2 = 0 . 91 for the most relevant features in the delineation of patches with similar height structure and water cycle conditions, respectively.

show abstract