Abstract. Wood is an essential component of rivers and plays a significant role in
ecology and morphology. It can be also considered a risk factor in rivers
due to its influence on erosion and flooding. Quantifying and characterizing
wood fluxes in rivers during floods would improve our understanding of the
key processes but are hindered by technical challenges. Among various
techniques for monitoring wood in rivers, streamside videography is a
powerful approach to quantify different characteristics of wood in rivers,
but past research has employed a manual approach that has many limitations.
In this work, we introduce new software for the automatic detection of wood
pieces in rivers. We apply different image analysis techniques such as
static and dynamic masks, object tracking, and object characterization to
minimize false positive and missed detections. To assess the software
performance, results are compared with manual detections of wood from the
same videos, which was a time-consuming process. Key parameters that affect
detection are assessed, including surface reflections, lighting conditions,
flow discharge, wood position relative to the camera, and the length of wood
pieces. Preliminary results had a 36 % rate of false positive detection,
primarily due to light reflection and water waves, but post-processing
reduced this rate to 15 %. The missed detection rate was 71 % of piece
numbers in the preliminary result, but post-processing reduced this error to
only 6.5 % of piece numbers and 13.5 % of volume. The high precision of
the software shows that it can be used to massively increase the quantity of
wood flux data in rivers around the world, potentially in real time. The
significant impact of post-processing indicates that it is necessary to
train the software in various situations (location, time span, weather
conditions) to ensure reliable results. Manual wood detections and
annotations for this work took over 150 labor hours. In comparison, the
presented software coupled with an appropriate post-processing step
performed the same task in real time (55 h) on a standard desktop computer.
Ultrasonographic Evaluation of Hip Joint in Newborns