首页 | 本学科首页   官方微博 | 高级检索  
     

基于数字图像处理的平原城镇河网分形特征研究
引用本文:陈 菁,刘昭辰,陈 丹,肖晨光,李建国,刘素芳. 基于数字图像处理的平原城镇河网分形特征研究[J]. 水资源与水工程学报, 2016, 27(6): 31-38
作者姓名:陈 菁  刘昭辰  陈 丹  肖晨光  李建国  刘素芳
作者单位:(河海大学 水利水电学院, 南京 210098)
基金项目:江苏省水利科技重点项目(2014006)
摘    要:城镇河网分形特征反映了城镇水系分布密集度、城镇河网水面率与水系连通程度,对预测城镇水文趋势、旱涝抵御能力、水系及地貌演变等具有指导意义。本文以苏州吴江桃源镇河网为研究对象,基于数字图像处理技术,采用盒计数法和Sandbox法对桃源镇河网的分形特征进行定量分析。结果表明:盒计数法和Sandbox法均较好地揭示了该地区河网分形特性,分维值分别为1.4085和1.7442。盒计数法12个标度下的栅格边长和非空盒数关系曲线呈外凸状,揭示了该地区分形规律随标度减小而减弱的特征;以5个随机点和形心为控制点,采用Sandbox法对河网骨架分形进行计算,揭示了不同点控制区域分形维数的差异性。Sandbox法分维较盒计数法精确,原因为盒计数法受到最小标度、分形体不规则性和边界效应的影响,因此可采用盒计数法的定性判别与Sandbox法的定量计算对城镇河网分维综合评价。桃源镇河网分形特征说明了该区域地貌平缓且土地被河网分割严重,其河网分维受制于河道离散、土地破碎与城镇化发展的矛盾。

关 键 词:数字图像处理  河网水系  分形维数  盒计数法  Sandbox法

Research on the fractal characteristics of town river networks based on the digital image processes
CHEN Jing,LIU Zhaochen,CHEN Dan,XIAO Chenguang,LI Jianguo,LIU Sufang. Research on the fractal characteristics of town river networks based on the digital image processes[J]. Journal of water resources and water engineering, 2016, 27(6): 31-38
Authors:CHEN Jing  LIU Zhaochen  CHEN Dan  XIAO Chenguang  LI Jianguo  LIU Sufang
Abstract:The fractal dimension characteristics of the town river system reflects the distribution density and activities of townriver system, water surface ratio and connectivity of river system, which favorsand provide guidance for the prediction of hydrologic trends, resisting ability of drought and flood, and river system and morphological evolution. In this research, the river system in Taoyuan Town, WujiangDistrict in Suzhou was chosen as the research object. The box-counting method and the sandbox method were used to quantitatively compute the fractal characterization of the town river networks, based on the digital image processing techniques. The results showed that, both the box-counting and the sandbox methodsrevealed the fractal characteristics of the river system and the fractal dimensions well, the fractal dimension values were 1.4085 and 1.7442, respectively. Forthe box-counting method attwelve scales, the relationshipcurves between gridside lengths and number of non-empty boxes wereconvex, revealing that the fractal rules of the river system in the studied site decreased as the scale decreased. For the sandbox method, the centroid of the fractal body and five randompoints were chosen as the control points for fractal dimensionanalysis of the river system and diversities of fractal dimensions among different controlled districts were revealed. The box-counting method was less precise than the sandbox method, because the box-counting method tended to be affected by the minimum scale, the irregularity of the fractal body and the border effects. When conducting the comprehensive assessment of fractal analysisfortown river system, iIt is preferable that the box-counting method can be used for qualitative analysis and the Sandbox method be used for quantitative calculation. The fractal features of river network in Taoyuan Town indicated that the landscapes were plain and the lands are severely segmented by the river network. The fractal dimensions of river system in this region are constrained by contradiction of the river network dispersity, land fragmentation and urbanization development.
Keywords:digital image processing   river network   fractal dimension   box-counting method   Sandbox method
本文献已被 CNKI 等数据库收录!
点击此处可从《水资源与水工程学报》浏览原始摘要信息
点击此处可从《水资源与水工程学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号