The Estimation Of Oil Thickness Based On Simplified Hapke Model

Now, with the development of offshore oil exploitation and oil transportation, the accidents about ship damage, offshore oil platform and oil storage and transportation equipment take place frequently. For the problem about the assessment of the disaster, how to estimate the oil spillage accurately has received more and more focus. With the development of remote sensing technology, hyperspectral technology has been used to segment oil spilling area and estimate of oil film thickness. This makes it possible to estimate the oil spillage volume.Hyperspectral image has many characteristics as one of remote sensing technologies, including high spectral resolution, abundant band information. The technology of unmixing mixed pixels in hyperspectral image has been an important topic in this field for almost twenty year. It decomposes the mixed pixels to obtain material compositions and abundance in mixed pixel with hyperspectral abundant band information. This paper is aimed at extracting water and oil endmember, and calculating the abundance of oil and water by unmixing technology to estimate oil film thickness.The advantages and disadvantages of the technology of linear mixing model and unlinear mixing model on unmixing mixed pixels are compared here firstly, and the simplified Hapke model is introuduced in, and is applied for laboratory oil mixed pixels and real hyperspectral oil spilling image to unmixing. The unmixing result proves the unmixing precision is higher than linear mixing model and the computation complexity is lower than unlinear mixing model. In addition, this paper obtains oil mixed pixels curves at different weathering time by AvaSpec spectrograph, so as to extract the stable wave region for decomposition. The spectral bands in this region can be enhanced during unmixing to contribute to accurate result, and proves the value and significance of research about weathering instability bands by analysis the unmixng results of laboratory oil mixed pixels and real hyperspectral oil spilling image.