Research Of Numerical Simulation And Remote Sensing Image Processing Of Spilled Oil

Nowdays, with the development of society and people’s living standard, oil resourcesare constantly being developed, the traffic volume of ships at sea is growing rapidly. It leadsto marine oil spill accidents happen frequently. Sea water pollution problems are becomingmore and more serious. Oil spill accidents happen in high frequency in marine pollution, itnot only results in great damage for production and life, but also causes damage to theenvironment. It is important to solve the oil spill problems quickly and efficiently, after theoil spill accident, it is urgent to simulate the oil spill on the sea. And the remote sensingimage digital work is also significant.In this paper, a simulation method of sea surface with oil slicks which is adapted to theshallow sea environment is presented, and a new method of image segmentation adapted tospill oil image is proposed. The main research in this article as follows:(1) First, this paper analyzes the characteristics of oil spill in the physical point. Asimulation method of sea surface with oil slicks which is adapted to the shallow seaenvironment is presented, which considers the water depth factor. The spectrum of the seasurface with oil slicks under define-depth water is calculated by TMA spectrum model andMarangoni oil theory model. According to the dispersion relation and Longuet-Higginsmodel, the displacement of the sea surface with oil slicks under define-depth water isevaluated. The properties of Kitaigorodskii depth function are analyzed.(2) Secondly, this paper preprocess oil spill image, across analyzing the preprocessingeffects of spilled oil images by using Lee filter, Kuan filter, Gamma-Map filter, enhancedLEE filter and enhanced Kuan filter algorithm. On this basic, the Gamma transform iscombined into the enhanced Lee filter, and thus improves the enhanced Lee filter. Thisalgorithm can suppress the noise and maintain the image edges and the texture informationwell.(3) Finally, a new method of image segmentation adapted to spilled oil image ispresented. It extracts regions of interest in image, and it’s the premise of image analyzing.This paper by improving the fuzzy C-means algorithm, a new algorithm is presented, thenew algorithm uses the image histogram as the initial value, this algorithm overcomes theshortcomings of the initial value sensitivity and segment spilled oil image effectively.