| 多维信息在改善沿海地区图像分类精度中的应用方法 |
| |
| 引用本文： | 赵寒冰,李加林.多维信息在改善沿海地区图像分类精度中的应用方法[J].海洋地质与第四纪地质,2004,24(2):125-129. |
| |
| 作者姓名： | 赵寒冰 李加林 |
| |
| 作者单位： | 1. 南京师范大学,地理科学学院,南京,210097
2. 南京师范大学,地理科学学院,南京,210097;宁波大学,人居环境研究所,宁波,315211 |
| |
| 基金项目： | 宁波市科技局项目(2002C10026)，浙江省教育厅项目(20030503) |
| |
| 摘 要： | 在某些专题研究中，由于研究目标、大气传输、下垫面及其他因素的限制，波谱分析的统计分类模式并不能完全满足研究需要。虽然多时相高精度图像可以提高分类精度，但常因其费用相当昂贵而难以实现。因此，如何在现有资料基础上，提取更多有效信息是亟需解决的问题。针对沿海地区特有的复杂地理现象，以TM图像为基础，通过量化手段进行最佳信息维数的确定与最佳信息维数组合的选择，探讨了信息维数对分类精度的影响。结果表明，在原有图像基础上，进行信息挖掘并重新组合可以得到能满足工作需要的分类精度，组合维数以6维最佳，更多的信息融合对精度影响不明显。
|
| 关 键 词： | 多维信息 遥感图像分类 分类精度分析 沿海地区 |
| 文章编号： | 0256-1492(2004)02-0125-05 |
| 修稿时间： | 2003年7月10日 |
THE METHOD TO IMPROVE THE IMAGE CLASSIFICATION PRECISION USING MULTI-DIMENSIONS INFORMATION IN THE SEA COAST REGION |
| |
| ZHAO Han-bing.THE METHOD TO IMPROVE THE IMAGE CLASSIFICATION PRECISION USING MULTI-DIMENSIONS INFORMATION IN THE SEA COAST REGION[J].Marine Geology & Quaternary Geology,2004,24(2):125-129. |
| |
| Authors: | ZHAO Han-bing |
| |
| Institution: | ZHAO Han-bing~ |
| |
| Abstract: | For some special research projects, the computer-aid image classification work based on spectral statistical method can not completely satisfy our research target because of the different directions of the research, atmosphere transfer, land use and other factors. But it is too expensive to buy all the related remote sensing images, so it is practical to manage to get more information from the existing data at hand. Based on the original image, we try to solve the problem of classification of seaboard region using a quantitative method. We collect message from the original image up to 75 dimensions and select 18 dimensions of the most benefit for classification. After trying all the possible combinations of the multi-dimensions information in three statistical functions, we found that the combination of six dimensions could give the best result of all the three methods and more dimensions gave no help for improving precision. Then we select one of the best results and apply it to the TM image of CIXI region in Zhejiang Province. The result of the final classification is far better than that by the original bands of the image. |
| |
| Keywords: | multidimensions information remote sensing classification precision's comparison |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
