346
May 2019
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
cannot be used in GNSS-denied environments (e.g. indoors).
To overcome these issues, this paper investigates the in-
tegration of information provided by an UWB positioning
system with image-based reconstruction to produce a metric
reconstruction. Furthermore, the orientation (with respect to
North-East directions) of the model is assessed thanks to the
use of inertial sensors included in the UWB devices. Results
of this integration are shown on two case studies in indoor
environments.
Digital Surface Model Generation from Multi-view Stereo Satellite
Imagery
Along with the improvement of the spatial resolution, mul-
tiple-view stereo satellite imagery has become a valuable
data-source for digital surface model generations. In 2016, a
multi-view stereo publicly benchmark of commercial satellite
imagery has been released by the John Hopkins University
Applied Physics Laboratory, USA. Motivated by this well-or-
ganized benchmark, we propose a pipeline to process multi-
view satellite imagery to accomplish digital surface models.
According to the tests differentiating the image view angles
and capture dates, the input images are pre-selected. We
apply the relative bias-compensated model for orientation,
and then generate the epipolar image pairs. The images are
matched by the modified tube-based Semi-Global Matching
method tSGM. Within the triangulation step, very dense
point clouds are produced, and are fused by a median filter
to generate the Digital Surface Model (DSM). A comparison
with the reference data shows, that the fused DSM generated
by our pipeline is accurate and robust.
Self-Supervised Convolutional Neural Networks for Plant Reconstruc-
tion Using Stereo Imagery
Stereo matching can provide complete and dense 3D re-
construction to study plant growth. Recently, high-quality
stereo matching results were achieved combining semi-global
matching with deep learning. However, due to a lack of suit-
able training data, this technique is not readily applicable
for plant reconstruction. We propose a self-supervised MC-
CNN scheme to calculate matching cost and test it for plant
reconstruction. The MC-CNN network is re-trained using the
initial matching results obtained from the standard MC-CNN
weights. For the experiment, close-range photogrammetric
imagery of an in-house plant is used. The results show that
the performance of self-supervised MC-CNN is superior to
the Census algorithm and comparable to MC-CNN trained by
a lidar point cloud. Another experiment is performed using
stereo imagery of a field beech tree. The proposed self-train-
ing strategy is tested and has proved capable of identifying
the drought condition of trees from the reconstructed leaves.
