Geography Reference
In-Depth Information
the characterisation of velocity distributions, transport
processes related tomixing, and discharge measurements.
Videography has also been successfully implemented
for the quantification of other processes in a variety of nat-
ural environments. Examples include the use of satellite
imagery to track atmospheric cloud movements (Leese
et al., 1971) or sea ice (Ninnis et al., 1986), and the use of
land-based cameras to quantify near shore wave patterns
(Holland et al., 1997) or the movement of subaque-
ous nearshore ripples (Becker et al., 2007). Within rivers,
efforts have also beenmade to track the velocities and con-
centrations of ice floes (Ferrick et al., 1992, Ettema et al.,
1997, Jasek et al., 2001, Bourgault, 2008) and the volume
of wood in transport and accumulations (Lyn et al., 2003,
Moulin and Pi egay, 2004, MacVicar et al., 2009).
Numerous advantages are driving the development of
videography for river monitoring. First, the images are
obtained remotely. Measurements are possible during
floods when it is too laborious or dangerous to attempt
them by other means, particularly when large floating
objects such as ice and wood are a concern (Fujita and
Hino, 2003, Hauet et al., 2008a). Velocities are also mea-
sured simultaneously over a large area, which is not
possible with other instruments. Second, the recording
of images is continuous, which means that relatively
rare events will be recorded. This advantage is especially
important in small basins where the time of concen-
tration is short. A third advantage is that videographic
systems can be connected to local and global networks.
Digitally recorded videos can be transmitted from remote
field sites to a central location for storage and analysis,
significantly reducing the cost of monitoring. Finally, the
proliferation of video monitoring for public security has
driven down both the cost of video monitoring systems
and the technical expertise required for their installation
and operation.
This chapter presents three case studies in which
videography was implemented to quantify stream dis-
charge, water velocity vectors, and the flux of floating
wood. These techniques rely on the detection of material
floating on the surface of the water, called the 'seeding', to
determine movement vectors. The estimation of the flux
of floating wood further requires that the types of seed-
ing material be distinguished using an object's secondary
properties such as color, linearity, and direction of move-
ment. This chapter is organized in four sections. General
considerations for a videographic system are presented in
Section 16.2 followed by a case study on stream gauging
using a mobile LSPIV system in Section 16.3, a case study
in which algorithms are developed to filter poor quality
LSPIV measurements in Section 16.4, and a case study
in which videography was used to measure wood flux
case study in Section 16.5. Further research questions and
applications are also discussed.
16.2 General considerations
16.2.1 Flowvisualisationand illumination
Image-based velocimetry requires high quality measure-
ments of object displacements. Laboratory investigations
of PIV have shown that it is necessary to consider the
concentration of particles, their size with respect with the
image processing parameters, and the anticipated particle
displacement to ensure accurate measurements (Adrian,
1991). Unfortunately, ideal conditions are rarely obtained
for field applications of videography. Illumination is not
constant due to solar effects, weather conditions, and
the time of day. As the angle of the sun changes, light
reflections, lens flare, shadows and glare can reduce the
quality of the recorded image (Kim, 2006, Hauet et al.,
2008b). Fog, rain and snow can reduce the visibility of the
surface. Luminosity is obviously greatly reduced at night.
In general, adverse lighting effects can be minimised with
a propitious selection of camera position. In the case stud-
ies presented in this chapter, mobile cameras were placed
on masts or tripods and fixed cameras were installed at
a high point on a bridge or bank to maximise the view
of the river and angle of the camera to the horizontal.
Lens shades were used and cameras were oriented away
from the sun, preferably towards the north or where trees
were likely to prevent direct sunlight from hitting the
lens. For nighttime illumination infrared assist options
were tested, including the installation of a high-powered
infrared light at one site. As part of this chapter, algo-
rithms are introduced that were used to reduce the errors
from illumination problems related to surface reflections,
shadows and other effects.
16.2.2 Recording
The capability to distinguish movement between two
image pairs is commensurate with the size of the images,
their resolution, seeding, and the time interval between
images. The studies presented in this chapter used com-
mercially available video cameras with image sizes less
than 720
576 pixels and measurement frequencies
between 5 and 30 Hz. The size of pixels in real coordinates
was highly variable because river widths varied from 10
to 70m and camera heights and angles were adjusted to
×
Search WWH ::




Custom Search