项目名称: 基于多尺度统计模型的SAR图像海洋表面油膜检测
项目编号: No.61201416
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 电子学与信息系统
项目作者: 金海燕
作者单位: 西安理工大学
项目金额: 24万元
中文摘要: 近几年兴起的多尺度几何分析方法主要致力于研究高维函数的最优表示,这类方法能非常好地"捕获"图像的奇异性信息- - 边缘。本项目提出把多尺度几何分析方法及其变换域统计模型用于SAR图像海洋表面油膜检测,充分发掘多尺度几何分析的潜力,有望得到更好的检测效果。本课题的主要研究内容包括:基于多尺度几何分析理论,建立变换域统计模型,进而探索能够实现SAR图像海洋表面油膜检测的具体过程,在验证其有效性的基础上,探讨如何综合利用各种多尺度几何分析的统计模型,进行组合应用。本项目的实现将为SAR图像海洋表面油膜检测的工程实现奠定必要的理论基础。
中文关键词: 图像处理;统计分析模型;优化算法;SAR图像;海洋表面油膜检测
英文摘要: Recent popular multiscale geometrical analysis methods mainly dedicate to optimized presentation for high-dimension function, which are good at capturing singular information of images - - edges. This project proposes to apply multiscale geometrical analysis methods and their translform domain statitical models to detect SAR image sea surface oil film, which can fully exert their potential and hopeful obtain better detection results. The main research contents of this project is as follows: build transform statitical models based on the theory of multiscale geometrical analysis,further more, explore the detail steps for realizing the detection of SAR image sea surface oil film; search for that how to synthetically utilize the statistical models of multiscale geometrical analysis to combinated application on the basis of testifying their validity. The realization of this project will establish necessary theory base for the engineering application of SAR image sea surface oil film detecion.
英文关键词: image processing;statistical analysis model;optimization algorithm;SAR image;detection for sea surface oil film