The Index Based On MODIS Oil Spill Research And Application

With the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor has been pursued and began transmitting data to the ground. Now, MODIS data has been used as one of the most widely remote sensing data sources. Marine oil spill on the environment is very harmful, particularly influence on the marine environment, marine organism et al. which is usually irreversible. So, it is essential to quickly and accurately detect the oil spill area. Due to offshore oil area influence by ocean currents, wind, et al, it has the feature-rapidly changes. The MODIS sensor transit time is four times in one day, so that it can be dynamic tracking, monitoring changes in oil spill area.The traditional division classification spill area as the computational complexity is often higher, it is not suitable for rapid segmentation and real-time detection about Spill area. In this paper, the design idea use various commonly index of remote sensing field for reference, an index oil spill (NDOI) has been proposed.2images of MODIS data (May2010Gulf of Mexico oil spill accident and June2011of Bohai oil spill accident) have been tested, experimental results demonstrate the effectiveness of a certain extent of the spill Index. In summary, the main jobs of this paper are as follows:Autocorrelation interpolation(1) pre-processing part:In this paper, using the autocorrelation interpolation method to remove stripe noises, the algorithm is not only simple and effective, but does not change the original MODIS data. This paper uses a traditional, rapid method to remove Bowtie effect use ephemeris to eliminate Bowtie effect. The main objective is to reduce some adverse factors of pre-processing part to provide more accurate data for subsequent work.(2) Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI):In this paper, detailed analysis of two main indices of NDVI and NDWI, including their principles, NDVI better identification of land, but oil is not pronounced, NDWI for land, water has a certain distinction effect, but is not ideal.(3) Oil Spill Index (NDOI) proposed:Based on the above problems, this paper MODIS spectral information of each band was analyzed and found that36MODIS bands, the evident difference between the top10bands curve around objects, MODIS images for visual interpretation and spatial analysis of spectral signatures, found the first of10bands have amount of information, the first one and second band have amount of information, especially rich. The features of seventh band most clearly the best. Using the above analysis, this paper proposes a spill index (NDOI), to classified oil water.