WThis report documents the algorithms, data processing and software for the video photogrammetry ("videogrammetry") system developed at Cranfield University. Cranfield's system has been used successfully since 1999 on a range of measurement projects. Videogrammetry typically uses two video cameras to film the motion of target objects in stereo, and then with suitable image processing and data analysis the targets' 3d trajectories are measured to good precision. The main features of the Cranfield systems are that it is based on consumer electronics devices (e.g. digital camcorders and PC's), and that it is designed as an experimental tool. Using consumer electronics provides good performance at low cost. Its experimental character means that an expert user is required, but does allow great flexibility. Current system performance derives from the image resolution of 1 mrad per pixel over a field of view 720 by 576 pixels, and a frame rate of 25 Hz. Two areas of work are described: (1) the mathematical models and algorithms used for calibration, position measurement and trajectory extraction, and (2) the software tools written to manipulate images and process the data. The model of the imaging system can be adapted for a wide range of applications, and is explicitly developed in this report for a single camera position and pose calibration, a two-camera system calibration, and a measurement sys- tem using two or more cameras (either the general non-linear case or a linear approximation). The image calibration (which converts image coordinates to geometrical angles of inclination and azimuth) is based on a 3rd order polyno- mial and achieves an accuracy equivalent to better than 1 pixel. Trajectories are obtained by either labelling targets or using kinematic rules. Two programs (AVI1 and mfitvid) have been written for the videogrammetry system and are described in outline (for a user and to support future development). Commercial software provides the more general functions required by the system. Experience with the videogrammetry system over a number of years gives confidence in its performance. Example results are provided to illustrate the type of measurements which are possible.

A technique for satellite formation flying modelling in LEO is applied at L2. Analytical solutions to the equations of motion of a hub satellite relative to L2 are used to define a halo reference orbit. An expression for the gravity gradient is obtained at the hub and the linearised equations of motion of the mirror satellites relative to the hub are derived. The relative motion model is implemented in Matlab/Simulink and evaluated for different initial conditions. The analytical solutions to the equations of relative motion are derived. These and other equations of motion are compared to the Satellite Tool Kit numerical orbit propagator.

We exploit recent advances in active high-resolution imaging through scattering media with ballistic photons. We derive the fundamental limits on the accuracy of the estimated parameters of a mathematical model that describes such an imaging scenario and compare the performance of ballistic and conventional imaging systems. This model is later used to derive optimal single-pixel statistical tests for detecting objects hidden in turbid media. To improve the detection rate of the aforementioned single-pixel detectors, we develop a multiscale algorithm based on the generalized likelihood ratio test framework. Moreover, considering the effect of diffraction, we derive a lower bound on the achievable spatial resolution of the proposed imaging systems. Furthermore, we present the first experimental ballistic scanner that directly takes advantage of novel adaptive sampling and reconstruction techniques.

Phase information has fundamental importance in many two-dimensional (2-D) signal processing problems. In this paper, we consider 2-D signals with random amplitude and a continuous deterministic phase. The signal is represented by a random amplitude polynomial-phase model. A computationally efficient estimation algorithm for the signal parameters is presented. The algorithm is based on the properties of the mean phase differencing operator, which is introduced and analyzed. Assuming that the signal is observed in additive white Gaussian noise and that the amplitude field is Gaussian as well, we derive the Cramer-Rao lower bound (CRB) on the error variance in jointly estimating the model parameters. The performance of the algorithm in the presence of additive white Gaussian noise is illustrated by numerical examples and compared with the CRB.

Paralytic shellfish poisoning (PSP) occurs nearly every year in the Gulf of Maine. In a study of dynamics of the causative organism, the toxic dinoflagellate Alexandrium sp., we conducted three surveys of the coastal and offshore waters of Gulf of Maine during the summer of 1998, sampling more than 200 stations during each cruise in June, July and August. Hydrographic data were collected and concentrations of phytoplankton chlorophyll, inorganic nutrients and densities of Alexandrium cells were measured in discrete water samples. The distributions of Alexandrium at the surface and in subsurface waters displayed maximum cell densities in the offshore waters of the Gulf on all three cruises. Highest cell densities in surface waters (ca. 5.5 x 10³ cells l¹) were observed in two broad patches: one in the Bay of Fundy and another in shelf and offshore waters of the central and eastern Gulf of Maine in association with the Eastern Maine Coastal Current. Highest subsurface densities of cells appeared to be associated with the frontal edges beyond the cold surface waters associated with the Eastern Maine Coastal Current. As the summer progressed, the highest surface densities of Alexandrium receded toward the eastern portions of the Gulf and the Bay of Fundy. We suggest that the offshore distributions of relatively high densities of Alexandrium are naturally occurring and can be related to inorganic nutrient fluxes, and to the ambient light field as it varies seasonally and vertically. Locations of high cell densities were described and interpreted using a nondimensional light-nutrient parameter, computed as the ratio of the depth of the 10% surface irradiance to the depth of 4 mMNO3 concentration. Possible mechanisms responsible for periodic development of PSP outbreaks in nearshore shellfish beds are discussed.

Gulf of Maine; Red Tides; Alexandrium; Nutrients; Phytoplankton; Hydrography

Coastal is an area that is influent by two processing factors that are marine and land dynamic process. In this area, occurs a complex dynamic process, which caused the relatively quick changes. The use of remote sensing data to study the process that occur in the area will get some helpful information to know the changes that happened, e.g. to study coastline dynamic. The need of this kind information is useful by countries that have long coastline and have so many islands just like Indonesia. This study use Landsat 7 data with ETM+ sensor, supported by Landsat 5 (TM), bathymetric map and tidal data from the study case area of Riau Archipelago especially Batam Island and surrounding. From this study can be summarize that the information on the coastline can be obtain in easy and accurate with combining band ratio of 4/2 and 5/2. The result has to be corrected with visual interpretation of color composite 543 RGB by visual editing to reduce errors from digital processing. The result also has to be corrected with tidal data because the information are obtain from different tidal time, which are usually different with the standard tidal data that are used to determine the coastline.

The project has achieved all of its objectives. Eighteen Landsat 7 ETM images were enhanced to highlight structures with topographic expression. These image products are available for download in several formats and map projections from the MapPlace. An image analysis framework has been added to the Exploration Assistant page of the MapPlace web site to allow basic image analysis processes to be performed on multispectral and hyperspectral images. Twenty Landsat 7 ETM images, five ASTER images and one AVIRIS image have been loaded into the system and are available for analysis. Five general analysis tools are included in this initial version of the site.

This project was designed to test the value and acceptance of providing image products and image analysis tools for use by the exploration community over the Internet. If warranted, additional enhanced Landsat imagery could be produced and added to the system. Addition types of multispectral and hyperspectral imagery can be added to the analysis system as well as additional analysis tools and more complete coverage of existing image types.

Geomorphic habitat mapping

The geomorphic habitats of 158 near-pristine estuaries were mapped in the Comparative Geomorphology of Estuaries Project by methods outlined in the Chapter 2. The maps comprise industry-standard ArcGIS files consisting of a base map of the estuary boundary, and vector layers showing the extent of subaerial, tidal and sub-tidal habitats. The maps are available for download in the OzEstuaries database ( where they exist alongside the maps of near-pristine and modified estuaries generated during the NLWRA (2002).

The near-pristine estuaries that were mapped in this project were chosen to reflect, as far as possible, the full range of geomorphic variability in Australia: A statistically significant number of the different types of estuaries (between 16 and 26) were chosen, and these were spread as far as practicable around the Australian coastline. These estuaries are located in all the coastal geomorphic regions of Australia (Harris et al. 2002). Within this context, state government preferences and the availability of suitable images and photographs were also taken into account. By ‘estuary types’, we refer to the major kinds of coastal waterways (i.e. wave- and tide-dominated estuaries and deltas, tidal creeks, strandplains and embayments) which experience differing amounts of wave, tide and river energy (Heap et al. 2001). By ‘coastal geomorphic regions’ (Harris et al. 2002) we refer to the North-west Coast, Gulf of Carpentaria, North-east Coast, South-east Coast and South-west Coast.

Near-pristine estuaries from Queensland and the Northern Territory comprised by far the largest proportion of the estuaries mapped during the project, accounting for 46% and 30% respectively. Significantly fewer estuaries were mapped in Western Australia (11%), Tasmania (9%), New South Wales (2%) and Victoria (<1%) style="font-weight: bold;">Remote sensing techniques and applications to near-pristine estuaries

As mentioned previously, near-pristine estuaries tend to be located along the most remote and inaccessible parts of the coastline, making them difficult to study. This is why remote sensing techniques (satellite-based or airborne sensors) and related applications (eCognitionTM) were scoped for their potential use in monitoring and gathering further information on the habitats and water quality of near-pristine estuaries. Remote sensing involves the collection of information about the earth’s surface, using sensors mounted on satellites or aeroplanes. The result of remote sensing is an image from which the spatial distribution of different landscape characteristics (e.g. mangrove forests and seagrass beds) can be mapped.

Remote sensing has been used for many years to map the coastal zone both for research and for management purposes. One relevant example was the use of aerial photography to map Australia’s coastal geomorphic habitats during the NLWRA. However, it is underscored in the remote sensing methods and applications portions of this report that more advanced satellite and /or airborne remote sensing instruments are now available, and these can be even more costeffective and objective than aerial photo interpretation. This is because the imagery can be collected at larger spatial scales (regional) and because computer software is now available that can help automate the process by which maps are made from the images. For example, the eCognitionTM software package is designed to produce maps from remote sensing images by grouping similar adjacent features (e.g. water body or forest patches) in the landscape. We present a case study from the near-pristine Wildman River in the Northern Territory (see Chapter 4) in which we demonstrate that maps made from remotely sensed images using eCognitionTM are comparable to or better than maps made using manual digitising.

The level of detail and number of physical properties that can be mapped using remote sensing imagery depend largely on the characteristics and availability ofthe remote sensing instruments. In a second case study on near-pristine estuaries in tropical northern Queensland (see Chapter 3), different types of imagery were investigated for their ability to differentiate between the dominant vegetation types in an estuarine environment. An automated mapping routine was then developed and applied to widely available satellite data (Landsat ETM) to produce maps of the Daintree River, Cooper Creek and Noah Creek. These maps were then compared to aerial photo interpretations. Although slightly different land-cover types were used, the results were comparable for the common vegetation classes. The advantage in using Landsat ETM (Enhanced Thematic Mapper) data in an automatic mapping routine over manual interpretation of aerial photographs is that it is fast and can cover large regions. The disadvantage is that user knowledge and experience cannot be applied at the same level of detail.

It can be summarised that remote sensing used in combination with eCognitionTM (or like-programs) can reduce the time and effort required for routine monitoring and mapping of Australia’s near-pristine estuaries, and allow environmental managers and researchers to regularly update habitat maps to monitor changes through time.

Language/Country:
English

Institution:
Cooperative Research Centre for Coastal Zone, Estuary and Waterway Management

Type :
PDF

Field :
GIS and Remote Sensing, Ecology

Years :
2006

Download :
Right click and choose save as this link
or click the beside pdf icon