典型火山岩型铀成矿区遥感图像分形特征研究

Study on fractal characteristics of remote sensing image in the typical volcanic uranium metallogenic areas

简要介绍遥感图像分形维数与多重分形谱的计算方法,利用分形方法研究相山和玉华山火山岩型铀成矿区遥感图像特征。研究发现,众多铀矿床分布的相山盆地与仅产出2个铀矿点的玉华山盆地相比,遥感图像不仅具有更大的纹理分形维数,而且多重分形谱明显偏向奇异性指数更小的一侧分布。分析铀成矿区内遥感图像分形维数大小和多重分形奇异性强弱与铀成矿的内在联系,认为遥感图像的分形维数和多重分形奇异性指数可用于指导火山岩型铀矿的区域预测。

Computing Methods of fractal dimension and multifractal spectrum about the remote sensing image are briefly introduced. The fractal method is used to study the characteristics of remote sensing images in Xiangshan and Yuhuashan volcanic uranium metallogenic areas in southern China. The research results indicate that the Xiangshan basin in which lots of volcanic uranium deposits occur,is of bigger fractal dimension based on remote sensing image texture than that of the Yuhuashan basin in which two uranium ore occurrences exist, and the multifractal spectrum in the Xiangshan basin obviously leans to less singularity index than in the Yuhuashan basin. The relation of the frac,tal dimension and multifractal singularity of remote sensing image to uranium metallogeny are discussed. The fractal dimension and multifractal singularity index of remote sensing image may be used to predict the volcanic uranium metallogenic areas.