Multi-sensor remote sensing expert systems for detecting anthropogenic hydrocarbon pollution

As a major aspect of marine pollution, oil release into the sea has become a common phenomenon and it can have serious biological and economic impacts. Accurate detection and forecast of oil spill in a timely manner would be beneficial to resource management for monitoring the marine environment. It is one of the most important applications for operational oceanography. It has been demonstrated that remote sensing is a tool that offers a non-destructive investigation method and has a significant added value to traditional methods. This research presents different satellite sensors and oil spill detectability under varying conditions. In particular, I concentrate on the use of automatic approaches to detect oil spills in different imagery (in passive and active remote sensing systems). I conclude with a discussion of suggestions for further research with respect to oil spill detection systems. In the first phase of my research, a study for examining the feasibility of passive remote sensing systems in order to detect oil spills pollutions has been done. The Landsat ETM+ images were used to study the oil spill in Gulf of Mexico. An attempt has been made to perform ratio operations to enhance the feature. The study concluded that the bands difference between 660 and 560 nm, division at 660 and 560 and division at 825 and 560 nm, normalized by 480 nm provide the best result. Multilayer perceptron neural network classifier is used in order to perform a pixel-based supervised classification. The result indicates the potential of Landsat ETM+ data in oil spill detection. In the second phase of my research, I have focused on active remote sensing systems for oil spill detection. Synthetic aperture radar (SAR) can provide valuable synoptic information about the position and size of the oil spill due to its wide area coverage and day/night, and all-weather capabilities. Detection of oil spills from SAR imagery can be divided into three steps: (1) Dark feature detection, (2) Computation and extraction of physical and geometrical features characterizing the dark feature, and (3) accurate discrimination between oil spills and look-alikes such as ice, internal waves, kelp beds, natural organics, jellyfish, algae, low wind speed areas (wind speed < 3 m/s) and rain cells. In fact, the extraction of the dark spots in the image is the first of three necessary steps, the other two being its characterization by using a set of features and the classification between oil spill and look-alike. Aside from the accuracy of the segmentation results, one of the most significant parameters for evaluating the performance in this context is the processing time which is necessary to provide the segmented image. As a part of this research, I present a new fast, robust and effective automated method for oil-spill monitoring. A new approach has been generated from the combination of Weibull Multiplicative Model and neural network techniques to differentiate between dark spots and the background. First, the filter created based on Weibull Multiplicative Model is applied to each sub-image. Second, the subimage is segmented by two different neural networks techniques (Pulsed Coupled Neural Networks and Multilayer Perceptron Neural Networks). As the last step, a very simple filtering process is used to eliminate the false targets. The proposed approaches were tested on 60 ENVISAT and ERS2 images which contained dark spots.