The Study On Fractal Characteristics Of Land Coverage Using Quickbird Date

In the last decade, an increasingly attention was paid on the method of extract information from meter-based high spatial remote sensing satellite imagery whose information is abundant in structure and geometry than the traditional remote sensing satellite imagery. The object-oriented approach was widely used in high resolution or texture imagery. The key technology of object-oriented image classification is that: the information being used to interpret the image doesn't exist in single pixel but in the interrelation of image objects. The classification technology based on object is different from the pure spectral information classification; the image object also includes many other features used to classify information such as shape, vein and interrelation and so on.The fractal theory is a very active branch of modem mathematics and non-lineal science. It has played an increasingly important role in image processing and analysis. Many reports have been made about the applications of fractal theory in the fields of natural image simulation, image texture analysis, pattern recognition and etc. The fractal dimension is a basic mathematical conception in the fractal theory and it is one of the most important factors in the applications of the fractal theory.The research works base on 15 image samples which is selected in five different LUC(Land use and Land Cover) category. The edge of sample image was attracted by using Canny arithmetic operators. With the MATLAB program, the binary images of the particles border were acquired, the numbers of a series of square blocks whose lengths were different pixel quantities to cover the binary image were counted, and the box-counting dimensions of these images were calculated according to the mathematics relationship of the pixel quantities and the numbers of square blocks. Research results validate that the edge of sample images has fractal feature, and also prove that image objects of different LUC can be discriminated depending on feature parameters of fractal dimension because the fractal dimensions of varieties of weed are different apparently. According to the characteristic of fractal being good at describing different image object, A classification method by using fractal dimension as image objects feature is provided on the Definiens Developer7.0 platform,which is viable and effect.