PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
April 2019
243
PHOTOGRAMME TR I C ENG I NE ER I NG & REMOT E SENS I NG
The official journal for imaging and geospatial information science and technology
April 2019 Volume 85 Number 4
FEATURES
PEER-REVIEWED ARTICLES
COLUMNS
This month we look at România.
ANNOUNCEMENTS
Join us in welcoming our newest members to
DEPARTMENTS
,
,
,
and
This paper introduces a novel adaptive cloud detection method based on the spectrum
and texture from super-pixels. The algorithm uses spectral thresholding on the spectral
attributes and angular second moment of the super-pixels which are then combined by a
region growing algorithm.
Kyle T. Peterson, Vasit Sagan, Paheding Sidike, Elizabeth A. Hasen-
mueller, John J. Sloan,
and
Jason H. Knouft
The approach presented here models the relationships between spectral reflectance and
water quality parameters. The water quality variable is predicted by machine learning (ML)
regression techniques. The resulting models are then combined using decision-level fusion.
Nicholas Forfinski-Sarkozi
and
Christopher Parrish
Article demonstrates a spaceborne data-fusion approach that has the potential to signifi-
cantly shrink the global nearshore data gap often referred to as the “white ribbon.” The
output bathymetric dataset generated by the approach agrees with high-resolution Fugro
LADS MK II bathymetric data to within an RMS difference of 1.1 m.
Michael Ying Yang, Wentong Liao, Xinbo Li, Yanpeng Cao,
and
Bodo
Rosenhahn
The paper proposes a double focal loss convolutional neural network framework
(DFL-CNN) for vehicle detection in aerial imagery. The paper also introduces the first
large-scale vehicle detection dataset ITCVD with ground truth annotations for all the
vehicles in the scene.
Qinghong Sheng, Rui Yang,
and
Hui Xiao
An approach that fuses three-dimensional points and four-dimensional lines in Plücker
coordinates to achieve integrated coordinate transformation for estimating relative
orientation. Experimental results show that the integrated adjustment solution of P-LPRO
is more accurate than the traditional methods.
Megan Ritelli
and
Ayman Habib
Quasi-Rigorous/Quasi-Simplified, a new pseudo-rigorous approach, is proposed for Lidar
system calibration. Unlike the Simplified approach, this approach can be used on any
type of terrain, can use non-parallel flight lines, and can incorporate control.
