A Study Of Marine Spilled Oil Detected Based On Orient Object

Spilt oil is one of the most serious marine environment disasters. Synthetic Aperture Radar (SAR) image has been wildly applied in ocean spilt oil detection. Traditional SAR image detection uses the full-range scan method, which has some disadvantages, includes low detection precision, time consuming and vast computer resources. The thesis puts forward a new method called orient-object spilt oil detection. It not only greatly increases the detection precision but also remarkably reduces the time of detection and compute resources it takes. This method lays good foundation for real-time ocean oil detection.In this paper, we pretreat the SAR image in the first step. Then we extract the suspected spilt oil objects and display its on the screen. After analyzing every suspected spilt oil object, it displays the detection result at last. In view that Neural Network has the abilities of self-organize, self-study and self-adapt, this paper combines texture analysis and Neural Network method for the classification of SAR images of spilt oil. Compared with the efficiency and accuracy of BP, Hopfiekld, ART-1 and RBF Neural Network Model, we choose the RBF Neural Network as the detection tool in the end. In the progress of texture analysis, four pixel texture parameters that are sensitive to SAR image of spilt oil are calculated by gray level difference statistics. The SAR image is classified by using the feature vector that is composed of the Gray Level Co-occurrence matrix features and gray of pixel.In the end of this paper, we contrast the two detection methods. The result shows that orient-object oil spill detection method yields satisfactory results.