关于湖泊轮廓变化趋势图像预测仿真研究.pdf

第33卷第06期
计算机仿真
2016年6月
文章编号:1006-9348(2016)06-0423-06
关于湖泊轮廓变化趋势图像预测仿真研究
宋秀秀,贾振红',杨杰2, Nikola Kasabov
(1.新疆大学信息科学与工程学院,新疆,新疆乌鲁木齐830046;
2.上海交通大学图像处理与模式识别研究所,上海200240;
3.新西兰奥克兰理工大学知识工程与开发研究所,新西兰奥克兰1020)
摘要:通过对图像变化的预测分析,研究了湖泊边缘的演变,预测得到下一步的湖泊轮廓图。针对影响湖泊演变的因素较
多、众多影响因素与湖泊轮廓演变存在非线性关系、传统方法仅依赖于数字对湖泊演变进行预测的问题,提出了更加直观化
的基于灰色预测的图像变化预测方法。以我国章江头木错湖为例,对其不同时期的 Landsat MSS/ TM/ETM+, CBERS CCD
遥感影像进行边界信息提取,获得湖泊轮廓图谱。在研究区域中间选择一个原点,并以原点做射线,得到预测所需的原始数
据。建立灰色预测模型,对所得数据进行预测分析,得到湖泊轮廓变化的预测图,并对预测结果进行验证,最终结果表明,改
进方法对湖泊轮廓变化趋势的预测是可行的并且有较高的预测精度。
关键词:轮廓演变;趋势预测;遥感影像;射线;灰色预测
中图分类号:TP393文献标识码:B
Simulation Study on Image Prediction of Lakes Outline Change
SONG XIU-XIU,JIA Zhen-hong, YANG Jie?, Nikola Kasabov
(1. School of Information Science and Engineering, Xinjiang University, Urumqi 830046, China;
2. Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Shanghai 200240, China
3. Knowledge Engineering and Discovery Research Institute, Auckland University
of Technology, Auckland 1020, New Zealand)
ABSTRACT Through analyzing the prediction of image changes and predicting the change of lake profile with time
the image of lakes outline later can be predicted. In view of the problem that the number of influencing factors is very
large, the output variable is a nonlinear function of the input variables and the traditional method only depends on the
numbers to forecast the lake evolution trend, we proposed a more intuitive method based on grey prediction of image
change. China Zhang jiang tou mu cuo lake was selected as the research object, the boundary information of the lake
was extracted from its remote sensing images of Landsat MSS/TM/ETM+, CBERS CCD in different periods, and the
outline of the lake was generated. Choose an origin in the middle of the study area and draw radial lines from the ori-
gin to obtain original data for the predicting. The data were loaded into gray prediction model to acquire the forecas
ting image of the lake, then prediction results was validated. The final results show that the research method has cer
tain rationality and validity
KEYWORDS: Profile evolution; Trend prediction; Remote sensing image; Ray; Grey prediction
图像完成了对小麦产量、地热资源、食品的大肠菌群的预
1引言
测。特别是在医学领域中,人们建立了基于医学图像的
图像变化预测已经应用在很多领域中。例如人们利用预测模型们。这些研究表明了图像应用于预测的有效性和
可行性,并且有越来越广泛的应用。利用图像进行预测,打
破了以前仅依赖于测量数据进行预测的局限,给预测开创了
基金项目:教育部促进与美大地区科研合作与高层次人才培养项目
个新思路。本文将采用基于遥感图像上的预测方法来研
(2014-2029
收稿日期:2015-08-02修回日期:2015-09-04
究溯泊、沙漠等生态系统的边缘演变及变化趋勢。
但对于生态系统演变趋势的预测基本都是对测量数据
万方数括