Collaborative Research on Numerical Methods for Image Processing

图像处理数值方法的合作研究

• 批准号: 9806413
• 负责人: Lothar Reichel
• 金额: $ 8.64万
• 依托单位: Kent State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1998
• 资助国家: 美国
• 起止时间: 1998-07-01 至 2001-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9806413&HistoricalAwards=false
• 关键词: Collaborative; Research; Numerical; Methods; Image

9806413 Reichel The numerical solution of image processing problems gives rise to very large nonsymmetric linear systems whose solutions are extremely sensitive to perturbations in the data. To avoid or at least contain the disastrous effects of such instability, regularization is required. Regularization yields slightly modified problems which are less sensitive to inaccuracies in the data. The aim of the research by the investigator and his colleague is to develop effective and efficient regularization methods for the solution of large nonsymmetric linear systems that arise in image processing applications, as well as to develop new algorithms for image compression. The new numerical methods will be tested on medical imaging data and will be applied to interventional magnetic resonance imaging in collaboration with the department for Biomedical Engineering at Case Western Reserve University.

9806413 Reichel图像处理问题的数值解产生了非常大的非对称线性系统，其解对数据中的扰动极其敏感。为了避免或至少控制这种不稳定的灾难性影响，需要进行正规化。正则化产生稍微修改的问题，这些问题对数据中的不准确不那么敏感。研究者和他的同事的研究目的是为解决图像处理应用中出现的大型非对称线性系统开发有效和高效的正则化方法，以及为图像压缩开发新的算法。新的数值方法将在医学成像数据上进行测试，并将与凯斯西储大学生物医学工程系合作应用于介入磁共振成像。