Automatic Recognition of Fold Landform Types Based on Spatial Structure Pattern Matching

  • CHEN Ying , 1 ,
  • LI Anbo , 1, 2, 3, 4, * ,
  • YAO Mengmeng 1 ,
  • LU Guonian 1, 2, 3, 4
Expand
  • 1. School of Geography Science, Nanjing Normal University, Nanjing 210023, China
  • 2. Key Laboratory of Virtual Geographic Environment (Nanjing Normal University), Ministry of Education, Nanjing 210023, China
  • 3. Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing 210023, China
  • 4. State Key Laboratory Cultivation Base of Geographical Environment Evolution (Jiangsu Province), Nanjing 210023, China;
*Corresponding author: LI Anbo, E-mail:

Received date: 2016-02-24

  Request revised date: 2016-04-15

  Online published: 2016-11-20

Copyright

《地球信息科学学报》编辑部 所有

Abstract

The automatic recognition of fold structure is one of the bases of tectonic interpretation, geomorphology classification and three-dimensional geological modeling. At present, most of the automatic recognition methods used for landform classification are based on the regular statistical unit. These methods, although effectively extract the characteristic landform by using image or terrain data, cannot recognize the tectonic landforms which combined the structural feature and topographical feature. As one of the most general tectonic landforms, fold landform has featured a symmetric repetitive spatial structure, which can be used to recognize the fold. To realize the automatic recognition of fold landform types, this research provides a method based on the spatial structure pattern matching. This method focuses on building scene models of fold structures by using Attributed Relational Graph (ARG) and identifying the fold landform types by defining different spatial structure patterns through the formal grammar. The implementation process is presented as follows. Firstly, extract the long strip scene that may contain the fold structure according to the principles used in choosing fold cores and section lines. Secondly, build and simplify the spatial structure model of the long strip scene by following the ARG approach. Thirdly, convert the ARG model into sentences, and classify the fold types with respect to different grammatical inferences of the sentences. If the sentences cannot be inferred by Anticline Grammar and Syncline Grammar, then it is not a fold. Fourthly, determine the fold landform types by checking whether the terrain containing the fold is a mountain or a valley. The result shows that the proposed method is capable for automatically recognizing the fold landform types in the northern Lushan area. It basically solves the problem in the auto-recognizing of fold landform types for mountainous area, and can be a supplementary reference to the traditional methods used for landform classification.

Cite this article

CHEN Ying , LI Anbo , YAO Mengmeng , LU Guonian . Automatic Recognition of Fold Landform Types Based on Spatial Structure Pattern Matching[J]. Journal of Geo-information Science, 2016 , 18(11) : 1500 -1512 . DOI: 10.3724/SP.J.1047.2016.01500

1 引言

地貌作为地表各自然地理要素相互作用最活跃的界面,是地理学研究的基础内容之一[1-3]。合理的地貌类型划分和特征提取是进行地貌研究的前提[4]。传统的地貌分类主要是在实地勘查获取地貌信息的基础上,通过手工勾绘地貌界线而成[4-5]。这种方法精细准确,但工作量大、历时长,且受限于制图人员的技术和认识,因此许多学者开始探讨地貌类型划分和特征提取的自动化方法[6]
目前,利用数字高程模型(Digital Elevation Model,DEM)进行半自动化-自动化的地貌识别是地貌分类的研究热点,其主要利用规则统计单元进行地貌形态自动划分[7]或根据地形特征线来分割地貌单元[8-9]等,基本解决了基于形态特征的地貌划分问题。此外,利用遥感数据进行半自动化地貌解译方法的研究取得一定进展,其方法主要是基于各遥感指数[10]或针对形状、纹理等解译特征[11-12]提取典型地物地貌,且相关研究已主要应用于火山[11]、沙漠[11]、岩溶[10]以及珊瑚海岸[12]地貌等影像特征明显的地貌类型。
然而,构造地貌因特有的空间结构特征,使其自动化识别提取有一定难度,目前相关研究较少。构造地貌是指通过内外地质动力的相互作用所形成的具有一定构造特征的地貌形式[13],其中,褶皱构造地貌与断裂构造地貌是2种最基本的构造地貌类型。断裂构造地貌主要包括断块山地、断层谷、断层崖等类型[14-16]。褶皱地貌主要包括背斜、向斜和穹隆等基本类型[14-16]表1)。
Tab.1 Basic types and characteristics of fold landforms

表1 基本褶皱地貌类型与特征[14-16]

基本褶皱地貌类型 特征
背斜 背斜山 对称重复,外新内老
背斜谷
向斜 向斜山 对称重复,外老内新
向斜谷
穹隆 等轴背斜隆起
目前,针对大型断裂构造地貌的半自动化识别研究有了一定的进展[17-19],但针对褶皱地貌而言,由于其存在以下3方面的特殊性,现有基于DEM或遥感影像的自动化地貌分类方法难以直接应用:(1)褶皱地貌作为一种地层组合结构,其划分不仅需要考虑地形信息,还需要考虑地质构造信息;(2)褶皱构造的主要特点在于其特有的地层“对称重复”空间结构,而常用的单元统计方法主要研究地表形态,对于空间结构问题并不适用;(3)在长期的外力作用下,褶皱可能进一步被侵蚀为逆地貌或再顺地貌[15],与其原生形态已大相径庭。传统的以地形起伏进行地貌类型划分的方法并不能作为褶皱地貌的划分标准。因此,本文从褶皱地貌的地层空间结构特征入手,主要针对背斜和向斜地貌进行识别。
褶皱反映了地壳发展及地貌演化过程,是构造地貌、构造地质等学科的主要研究对象,其自动化识别是构造解译、地貌划分、三维建模等工作的基础。为此,本文拟采用地形地质图、DEM等数据,基于空间结构模式匹配方法解决褶皱地貌的自动化识别问题。

2 研究思路

虽然褶皱构造形态复杂、类型多样,且没有具体的边界,但其本质为一组受力弯曲的地层[20],在剥蚀后呈现出核部中心对称、模式化的特定地层组合结构。为此,可在不同褶皱类型的空间结构模式定义基础上,先通过空间结构模式匹配的方法来实现褶皱构造识别,再结合地形信息确定褶皱构造地貌类型。具体实现步骤如图1所示。
Fig.1 Framework of automatically recognition of fold landform types

图1 褶皱地貌类型自动识别框架

(1)为提高识别准确度及效率,先对可能形成褶皱的区域进行裁取,生成识别条带场景用于建立模型及判别地形。
(2)基于识别条带绘制邻接属性关系图(Attributed Relational Graph,ARG),建立识别条带场景的ARG模型。为了与褶皱模式进行匹配,需要剔除邻接ARG模型中的后期侵入岩地层对象、第四纪覆盖地层对象以及无关方向上的对象。
(3)给定满足褶皱模式的形式文法定义,将化简后的场景邻接ARG模型表达为2个句子。根据句子能否由模式文法推导出,来判定该场景是否具有褶皱的“对称重复”结构,且依据推导出的最长子句来确定褶皱的范围。通过本环节工作,可根据空间结构模式匹配出背斜、向斜等基本褶皱类型。
(4)根据褶皱构造范围内的局部地形信息,将向斜与背斜构造进一步划分为向斜山、向斜谷、背斜山与背斜谷4个基本褶皱构造地貌类型。

3 研究方法

3.1 识别条带的提取

褶皱结构识别的前提是找到褶皱可能存在的位置,继而锁定可能组成褶皱的地层。可基于2个常用的原则来实现褶皱识别目标的锁定:① 核部地层选取原则(表2)。在地质图上,进行褶皱的人工判别,通常是先找到一个较新或较老的地层(即核部地层),再观察其左右地层是否存在对称分布的现象[14],因此寻找核部地层是褶皱识别的前提;②剖面线选取原则。横切褶皱所绘制的剖面图能有效地反映褶皱横剖面的形态和地层组合特征[20],其剖面线一般选定地层出露较全的部位,并尽量垂直区内构造和地层走向[21],因此垂直于褶皱走向的剖面线所穿过的一系列地层最可能反映该褶皱的结构。
基于上述原则,本文先获取可能的核部地层,并裁取垂直于核部地层走向的一组地层作为识别条带。识别条带所表达的场景,反映了条带所处区域地理场景的结构特征。对于存在“对称重复”空间结构的条带,可认为该区域存在褶皱构造。从而,可通过识别条带的提取,将整个研究区域内的褶皱识别问题,转化为有限识别条带内的褶皱识别问题。这样通过消除复杂地层多边形的影响,既降低了场景表达的难度,又可提高匹配的效率和准确度。
Tab.2 Rules for identifying core stratums

表2 核部地层识别规则

自然语言描述 形式化语言描述 示例
地层S相对所有邻接地层S年代较老或较新 If IsOld(S,S) = Ture
Or IsNew(S,S) = Ture
Then IsCore(S) = Ture
地层S基本被某邻接地层S环绕而内部不存在洞(岛) If IsSurrounded
(S,S) = Ture
Then IsCore(S) = Ture
地层S的邻接地层S集合中存在地层符号属性一致的地层 If HasSameAge(S)
Then IsCore(S) = Ture
识别条带的具体生成步骤为:① 在忽略侵入岩与第四纪覆盖地层的情况下,基于核部地层识别规则(表2)筛选出可能的核部地层;② 基于横切剖面方法,垂直于核部地层的走向绘制剖面线,裁取剖面线一定缓冲域内的地层要素集合作为识别条带。

3.2 基于ARG图的条带场景建模与化简

3.2.1 场景模型
若将识别条带视为一个空间场景,则褶皱识别的问题转化为条带空间场景的建模及基于褶皱场景结构的模式匹配问题。空间场景是空间对象及对象间关系的综合[22],其建模的关键是对场景中的对象和关系进行形式化表达[23-24]。相关研究中,较为代表性的有“逐步转变”模型[25]、SIMDTC算法[26]和ARG[27]等。其中,ARG图是将统计和结构2种方法结合起来的一种场景描述方法,既具有图的直观结构,又包含属性特征信息,描述能力较强[28]。为此,本文选用邻接ARG方法对条带场景逐个建模,进行形式化表达。
在ARG中,G=(V,E)中[24],V为顶点集合,代表对象;E为有向边的集合,代表对象之间的空间关系[29],空间关系模型众多,根据实际情况选取关系模型及含义(表3)。
Tab.3 Relational models used in ARG

表3 ARG中采用的关系模型

关系 描述 备注
T T1(相邻),T2(相离) 四交模型[31]
A A1(北),A2(北东),A3(东北),A4(东),A5(东南),A6(南东),A7(南),A8(南西),A9(西南),A10(西),A11(西北),A12(北西) 十二方向的方位模型[32]
D 地层符号在地层年代中的位置的代数差运算,D>0
本文中,V为地层对象,E为邻接地层间的时空关系,主要包括空间拓扑关系、时间距离关系和空间方位关系,即E={T,A,D},各构模对象含义如下:
(1)地层对象(用V表示):以质心点代替条带上裁取的地层面要素[30],作为顶点对象存储,其属性为相应地层的属性。
(2)地层间空间拓扑关系(用T表示):地层要素间仅包含相邻和相离2种拓扑关系,存在边连接的为相邻对象,不存在边连接的即为相离对象。
(3)地层间空间方位关系(用A表示):用有向边来表示相邻两地层间的方位关系,边由老地层指向新地层,代表新地层相对其相邻老地层的方位,间接表达了地层年代由老变新的方向。
(4)地层间时间距离关系(用D表示):以地层年代表序列为依据,地层对象的值为地层符号在序列中的位置。有向边终点的地层对象与起点的地层对象的值做代数差,结果作为2个地层对象间的时间距离。
3.2.2 场景ARG建模方法
条带场景由一组地层要素构成,其中包含了地层对象的信息、地层间的关系以及地层组合的结构等,对这些信息进行计算和提取,建立场景的邻接ARG模型。由场景条带建立邻接ARG模型主要包括以下5个环节:
(1)计算场景条带上每个地层要素的质心点,存储为地层对象,将地层要素的属性传给其质心的地层对象;
(2)根据地层年代表,用地层符号Age在地层年代表中的索引AgeId,来代表地层对象的新老,记入属性;
(3)遍历找到每个地层在场景条带上的邻接地层,建立邻接表。邻接表中对象间的空间拓扑关系都为T1;
(4)在邻接表中比较相邻地层对象的新老,以相对老的地层作为起点对象,相对新的地层作为终点对象,生成有向边。计算由起点指向终点的向量的方位角,作为2个地层间的空间方位关系A,存入有向边的属性表;
(5)将有向边终点对象与起点对象的索引值AgeId做差,结果作为2个地层对象的时间距离关系,存入有向边的属性表。
条带场景邻接ARG建模效果如图2所示。其中,圆表示地层对象,其属性包括地层编号、地层符号等;有向边的属性表示了空间邻接对象间的空间方位关系与时间距离关系。
Fig.2 Example of a scene modeling

图2 场景建模示例

3.2.3 场景ARG模型的化简
褶皱形成后,可能遭受后期地质作用的影响,从而在场景ARG模型中出现干扰对象,给基于ARG模型的褶皱结构自动识别带来一定难度。为此,要对相关干扰对象进行化简,并删除无关方向上的对象,将邻接ARG模型还原至原始褶皱地层结构。针对邻接ARG模型的3种主要干扰情况,化简方案如下:
(1)对于邻接地层地质年代属性相同的情况,直接合并处理。
(2)侵入单元与第四纪覆盖等干扰对象的化简,主要是剔除后期侵入单元对象和第四纪覆盖对象,并对其邻接地层进行合并、删除、更新等处理。具体情况如表4所示。
(3)无关方向上的对象。褶皱是垂直走向上一组呈两翼对称关系的地层,对于非垂直方向上的对象及其邻接关系可进行剔除处理。

3.3 褶皱构造的空间结构模式定义

完整出露的褶皱构造,其地层对象在垂向上具有“对称重复”的特定空间结构,且因核部、翼部间的新老关系不同,进一步表现为背斜与向斜2种不同的空间结构模式[33]。基于上述ARG场景建模方法,可建立背斜与向斜2种典型褶皱类型的邻接ARG场景概念模型,其模型如图3所示。其中,An与Am分别表示褶皱两翼由左至右、由右至左的空间方位关系。
Fig.3 Conceptual model of fold scene based on ARG

图3 基于ARG的褶皱场景概念模型

结构模式匹配的研究多见于制图综合中的路网模式匹配[33-34],采用既定的结构模式进行匹配,以保证综合前后的地图路网的合理性。由上述概念模型可知,褶皱中的地层组合没有固定长度,表现为一种递归的结构模式,难以用邻接ARG的结构模式方法表达其结构模式,而仅用于概念描述。为此,本文考虑采用基于文法定义匹配句子的方法,进行褶皱结构模式的定义和匹配研究。形式文法一方面表现出动态递归的特性,另外也可通过文法定义来约束句子结构[35-36]
因此,由背斜与向斜的场景概念模型可总结得到2条文法描述:① 地层组合关于核部地层对称;② 褶皱由核部至两翼,地层的新老关系变化单调。对于背斜和向斜2种邻接ARG概念模型,分别有如下文法定义[37]
3.3.1 背斜模式文法定义
定义1:地层组合模式文法G(S)={VS,VAges,S,PS},其中:
(1)VS是文法非终结符的集合,是一个非空有限集,VS ={S};
(2)VAges是文法的终结符,是一个非空有限集,是进行匹配的基本单位。集合 a 3 , , a n } 为地层年代表的映射字符串;n为地层年代表内地层符号的个数。如图4,示例地层年代表内的18个地层符号分别与字母 { a , b , c , , p 一一映射。
(3)S是文法推导的初始符号;
(4)PS是有限个文法产生式的集合,已知褶皱出露的地层组合呈中心对称形式,其定义:
PS:
S→a1Sa1 | a2Sa2| a3Sa3| | anSan
S→a1aia1 | a2aia2| | anaian
其中,i=1,2,3,…, n。其最短结构是由3个字符组成的中心对称字符串,中心的核部地层可能为任意字母ai,其中i=1,2,3,…,n
定义2:新老关系方位模式文法G(A)={VA, VDirectionS,A,PA},其中:
(1)VA是文法非终结符的集合,是一个非空有限集,VA={A};
(2)VDirectionS是文法的终结符,是一个非空有限集。VDirectionS={l,r},该集合表示了由核部到翼部地层对象由老到新的方位关系。取核部向右地层变新的方向为右,记为r,则反向为左,记为l;
(3)A是文法推导的初始符号;
(4)PA是有限个文法产生式的集合,其定义:PA: A→lr | lAr。
3.3.2 向斜模式文法的定义
地层组合模式文法G(S)={VS,VAges,S,PS},与背斜模式相同。
定义3:新老方位关系模式文法G(A)={VA, VDirectionS,A,PA},与背斜模式类似。其中:
(1)文法非终结符集合VA、文法的终结符集合VDirectionS与初始符号A的定义与背斜模式相同;
(2)PA是有限个文法产生式的集合,其定义: PA : A→rl | rAl。
Fig.4 Mapping stratigraphic units to a table

图4 地层年代映射表

3.4 褶皱构造模式匹配

基于上述褶皱模式的文法定义,将褶皱结构模式匹配,转换为文法与句子的模式匹配。对于识别单元的场景模型化简得到的邻接ARG模型,可进一步表达为2个句子;然后,通过判别句子能否由模式文法推导出来,实现针对不同褶皱构造模式的匹配识别。具体流程如图5所示。
Fig.5 Pattern matching process of fold structure

图5 褶皱构造模式的匹配流程

(1)将邻接ARG模型表达为地层组合句子ω与新老关系句子κ,并明确其核部地层Score。如表4示例模型与其生成的句子,用下划线标识核部地层;
(2)第一次推导,取Score及其左右邻接地层Sleft1、Sright1组成的句子ω1,Score与Sleft1、Sright1的新老关系方位组成句子κ1,判断ω1是否可以通过文法G(S)推导得到,κ1是否可以通过文法G(A)推导得到;
(3)若句子ω1、κ1分别可以通过文法G(S)、G(A)推导得到,则记录当前匹配的地层对象组合SList;若至少有一个句子不能由对应文法推导得到,则当前场景匹配结果为“不是褶皱”,匹配结束。表4的示例1存在可成功推导2个文法的最小子句,示例2则不可以;
(4)进一步,取组合SList的左右邻接地层组成新的句子,判断其地层对象组合句子及新老关系方位句子是否可以通过文法G(S)、G(A)推导得到;若2个句子都可以推导得到,记录当前匹配的新的地层对象组合SList,重复步骤(4),反之则执行步骤(5);
(5)保留匹配成功的最长子句,匹配结束。若满足定义1与定义2则为“背斜模式”,满足定义1与定义3则为“向斜模式”。如表5示例一推导得到id编号为“4、2、0、1、3”的最长子句,满足向斜模式。
Tab.4 Simplification scheme of adjacent ARG model

表4 场景邻接ARG模型的化简方案

情景 邻接ARG示例 示例描述 化简方案
化简前 化简后
化简对象位于属性不同的2个对象之间 化简对象B与对象A、C邻接,A与C属性不同 剔除B,剔除关系AB、BC,更新对象A和C的坐标,以及关系AC
化简对象位于属性相同的2个对象之间 化简对象B与对象A、C邻接,A与C属性相同同 剔除B,剔除关系AB、BC,合并对象A和C,更新与A、C关联的关系
化简对象有多于2个邻接对象 化简对象E与对象A、B、C、D、F与G相邻,A与B属性相同且方位与地层走向一致,C与D,F与G情况同上 剔除E,剔除关系AE、BE、CE、DE、FE和GE,合并A与B、C与D、F与G,更新与A、B、C、D、F、G关联的关系
Tab.5 Example of the pattern matching of fold structure

表5 褶皱构造模式匹配示例

示例 句子 待推导子句 推导过程 结果
示例1 jdabcbadg
llrrllrr
bcb
rl
S→bcb
A→rl
201
向斜
abcba
rrll
S→aSa→abcba
A→rAl→rrll
42013
向斜
dabcbad
lrrllr
S→dSd→daSad→dabcbad
不符合文法G(A)
输出42013
向斜
示例2 edcba
llll
dcb
ll
不符合文法G(S)
不符合文法G(A)
输出
非褶皱

3.5 褶皱形态识别

褶皱地貌类型的确定既需考虑成因,又需考虑形态[38]。确定褶皱构造类型之后,根据褶皱范围内的地形走向,进一步判别出山或谷的地貌形态。
本文基于识别场景的剖面线,通过比较褶皱内核部地层、最左地层与最右地层中心点的海拔,来判别地形。设识别得到核部地层core、最左地层leftn与最右地层rightn对应的海拔位置分别为C、L和R。不同情况的判别规则为:① 若C相对于L、R地势较低,则认为褶皱构造范围内地形以谷为主,如图6中的F1;② 若C相对于L、R地势较高,则认为褶皱构造范围内地形以山为主,如图6中的F2;③ 若C位于L、R高程之间,则褶皱构造范围内地形是坡或山谷并存,此时,判定地形以山为主。判别算法阐述如下:
(1)将剖面线l与褶皱构造(图6中的F1或F2)地层对象相交,计算剖面线在所有地层面要素的中点集P={lefti,lefti-1 left2,left1,core,right1 righti-1,righti};
(2)取P中代表最左、核部和最右的点{lefti,core,righti },在DEM上分别求取高程,记为{L,C,R};
(3)比较3点的海拔变化趋势,观察L-C与R-C的正负。若L-C与R-C均为正,则认为核部低于左右,形态近似谷;若L-C与R-C均为负,则认为核部高于左右,形态近似山;若L-C与R-C一正一负,则认为形态为坡或山谷并存,识别为山。
上述工作完成后,可以识别出褶皱构造范围内的地形,继而将褶皱构造识别结果与地形判别的结果综合,确定褶皱地貌类型,并标注褶皱地貌单元的大致范围。
褶皱地貌单元范围的标注,主要分为2种情况进行处理:① 待标注褶皱独立,与其余褶皱地层不重叠时,则以推导最长子句的范围为短轴,以核部地层走向上长度为长轴,划定褶皱单元;② 待标注褶皱与其余褶皱地层有重叠时,重叠的翼部地层为相邻的向斜与背斜共用[16];但为保证褶皱地貌单元的完整性,需要对重叠部分进行划分。则取翼部重复地层的中点均分翼部,将相邻的背斜向斜划分为不重复的单元,并依此确定短轴长度。
Fig.6 Classification of terrain with fold structure

图6 褶皱构造地形判别

4 实验与分析

4.1 实验区域和实验数据

庐山是典型的断块山,受地壳上升运动和南北水平运动的影响,呈东北-西南向伸展,山体呈椭圆形,长约25 km,宽约10 km。该区地层复杂,跨越年代大,岩浆岩、变质岩、沉积岩均有较多分布。以九奇峰-仰天坪为界,南部以断裂构造地貌为主,北部以褶曲构造地貌为主[39]。庐山北部表现有类型丰富的原生地貌和次生地貌,主要受复背斜构造控制,以南华系下统莲沱组地层为主,形成“三背两向,五岭四谷”的形态[40]。本文实验区选定庐山五老峰以北、仙人洞以东的主要山地地区。
实验采用的是1:50 000的庐山北部地区的地质体面数据(图7)与DEM数据,包括震旦纪(Z),寒武纪(€),奥陶纪(O),志留纪(S)地层,并有穿插覆盖了白垩纪(K)侵入岩单元与第四纪(Q)地层覆盖物。实验数据的属性以行业标准《数字化地质图图层及属性文件格式》(DZ/T 0197-1997)[41]为准,不符合标准的数据需要进行预处理。
依据实验数据,可以根据地层年代表[42]先得到地层年代符号对应的字符。除去侵入单元与第四纪地层,将莲沱组(Z_1l)至茅山组下段(S_3m^1)共18个地层的位置{1,2,3, ,18},依次赋值为{a,b, c, ,r}作为地层句子的终结符。
Fig.7 Experimental data

图7 实验数据(庐山北部区域地形地质图)

4.2 实验结果与分析

基于上述方法,在VS.Net 2010平台下,利用C#语言和DotSpatial 1.7组件,开发了原型系统。系统提供了核部地层选取、识别条带提取、条带场景建模等模块,实现了褶皱地貌类型自动化识别的功能,并自动划定褶皱地貌单元。相关实验结果与分析如下。
4.2.1 场景提取
基于实验数据,根据核部地层选取原则,得到36个核部地层,在核部地层垂向上做识别条带(图8),并记录核部地层的编号和走向。选取的核部地层主要为本区域较老的莲沱组(Z_1l)地层,与较新的坟头组(S_2f)地层,大部分识别条带走向为南东走向。
Fig.8 Result of scene extraction

图8 场景提取结果

4.2.2 场景邻接ARG建模与简化
对提取的场景条带进行建模。首先,计算每个地层对象的质心点绘制成点图层,并将地层属性传递给点;其次,计算两两相邻的地层对象间的空间方位关系和时间距离关系,绘制线图层,并将地层关系存储为线属性。示例条带内包含id编号为143至159的17个地层对象,其邻接ARG图与邻接ARG场景模型展示如图9(a),其中着重表示的9个对象为保留地层对象,id编号为153的地层为核部地层。化简后效果和邻接ARG模型如图9(b)。
4.2.3 褶皱结构自动识别结果
对于化简后的识别场景模型,用褶皱结构模式进行匹配。图10为示例场景模型匹配的结果,红色范围为符合褶皱结构模式的部分,并且可以判断为向斜褶皱。
对全图进行模式匹配,可以得到符合褶皱模式的条带部分有7个,得到的褶皱主要为震旦纪莲沱组的地层与寒武纪乌石门灰岩地层2个年代。匹配的条带场景中的片段展示如图11所示,其化简后ARG模型对应的句子与匹配的子句如表6所示。
Tab.6 Fold structure′s matching result

表6 褶皱构造匹配结果

编号 句子 匹配部分 结果
1 cbcdcblrrll bcdcbrrll 向斜
2 cbcdbcdcbclrrlrrllr cbclr 背斜
3 fecbcdbcrllrrlr cbclr 背斜
4 gcbcdcbcbcllrrllrlr bcdcbrrll 向斜
5 fgdcbcbcilrlllrlrrr cbclr 背斜
6 cbghjhijilrrrlrrl hjhrl 向斜
7 baghjhijilrrrlrlr ijirl 向斜
4.2.4 识别结果与分析
叠加地形形态的判别之后,判别得到基本褶皱地貌类型的结果(图12)。沿核部地层走向进行褶皱地貌单元的简单标注。
针对研究区域,海会幅与庐山幅的地质说明书[42]中明确研究区域内褶皱有7个(图13),本方法识别出了其中6个,大部分为震旦纪时代以莲沱组为主的地层出露。褶皱天花井背斜的褶皱对称结构因受北东向脆-韧性变形带切割位移而被破坏,未能正确识别。
另外,资料也可能存在遗漏或尺度问题,可能有一些较小的褶皱并未被资料记录。如图12东北部,在向斜山的西侧还识别出1个向斜谷,海会幅地质图说明书[42]中未记载。核对地层信息与产状信息后,推测该向斜谷可能与马祖山向斜山为同一褶皱在地质图上出露的不同部分。
Fig.9 Simplification of adjacent ARG model

图9 邻接ARG场景模型的化简

Fig.10 Example of fold pattern matching

图10 褶皱构造模式匹配示例

Fig.11 Identification result

图11 褶皱构造场景条带识别结果

Fig.12 Marking the fold landforms

图12 褶皱地貌类型标注

Fig.13 Fold axis in the experimental area

图13 实验区褶皱轴迹[42]

5 讨论

(1)在提取识别场景条带时,可能会出现选取不到核部地层或岩层属性中不包含走向信息的情形。相关处理方法如下:①搜索不到核部地层时,应基于图幅内所有地层多边形的走向,制作走向玫瑰图,再选取频数最大的角度,等间距裁取图幅生成识别条带,继而利用上述原则在识别条带内部寻找核部地层。②当属性中不包含岩层走向时,可计算多边形走向来替代地层走向。
(2)地貌单元的提取涉及到单元界线的划分,本文方法仅简单划定了褶皱地貌单元的范围。对于褶皱重叠部分,主要以重叠翼部的中点进行划分。此外,相邻的山谷也可能有重叠,亦可作为划分标准。因山谷相邻,彼此过渡,可取山峰最高点与谷地最低点海拔差的中间点作为分界,划分山谷单元。由于篇幅问题,本文仅阐述了大致思路,并褶皱地貌单元进行简单标注。褶皱地貌单元边界的自动化划分,今后将专门进行研究;
(3)本文方法主要适用于基岩山地区域褶皱类型的自动化识别,且着重针对向斜、背斜2种褶皱构造类型进行了研究。对于实质为一等轴背斜隆起的穹隆构造,理论上本文方法也同样适用。

6 结语

本文以基于ARG的褶皱场景空间结构建模方法为切入点,在研究利用形式文法进行不同褶皱空间结构模式定义的基础上,着重形成了一种基于空间结构模式匹配的褶皱构造自动识别方法,进而有效地支持了褶皱地貌的自动识别。本文方法可以有效地识别提取出庐山北部山区对称、完整的褶皱构造地貌单元,基本解决了构造地貌类型的自动化识别问题,是对传统地貌形态划分方法的有效补充。主要研究结论包括:
(1)基于空间结构进行构造地貌类型识别的方法,给具有空间结构特征的地貌类型识别提供了新思路,对其他形态结构地貌类型的自动化识别研究有一定的借鉴意义;
(2)基于邻接ARG建模方法,有效地实现了对地理场景的建模与匹配识别;
(3)基于形式文法,实现对不同褶皱类型的空间结构模式的定义。在将变长、线性的场景ARG模型转化为句子的前提下,利用句子的模式文法推导结果进行判别,有效实现了不同褶皱类型的匹配识别,形成了一种非静态空间结构模式匹配问题的可选方案。

The authors have declared that no competing interests exist.

[1]
Hammond E H.Analysis of properties in landform geography: an application to broad-scale landform mapping[J]. Annals of Association of American Geographers, 1964,54:11-19.First page of article

DOI

[2]
卢金发. 黄河中游流域地貌形态对流域产沙量的影响[J].地理研究,2002,21(2):171-178.在黄河中游地区 ,选择了 5 0多个面积约 5 0 0~ 2 5 0 0平方公里的水文测站流域 ,分别代表 6种不同自然地理类型 ,在流域沟壑密度、沟间地坡度小于 15°面积百分比等地貌形态指标量计的基础上 ,进行了流域产沙量与地貌形态指标相关分析。结果表明 ,对于不同类型流域 ,流域产沙量随流域地貌的变化遵循不同的响应规律 ,而且视流域其它下垫面环境条件的均一程度 ,其相关程度和响应速率各不相同。受地面物质、植被、地貌发育阶段等流域其它下垫面环境条件的制约 ,除沟壑密度外 ,流域产沙量与流域地貌形态的关系都没有人们以前所预期的好。

DOI

[ Lu J F.Effect of basin morphology on sediment yield in the middle reaches of the Yellow River[J]. Geographical Research, 2002,21(2):171-178. ]

[3]
Stallins J A.Geomorphology and ecology: unifying themes for complex systems in biogeomorphology[J]. Geomorphology, 2006,77:207-216.The interaction of geomorphic and ecologic landscape components has been largely conceptualized as independent. In one direction, geomorphic processes and landforms shape the distribution of biota. Conversely, in the other direction, biota modify geomorphic processes and landforms. Increasingly, the interactions between geomorphic and ecological components are more circular and developmentally intertwined. In this paper, I integrate these two independent perspectives within the framework of complexity theory. I outline four themes that characterize complex systems in biogeomorphology: multiple causality and the concept of recursivity, the influence of organisms that function as ecosystem engineers, the expression of an ecological topology, and ecological memory. Implicit in all of these themes is the recognition that biogeomorphic systems are open and path dependent. They may exhibit a range of assembly states, from self-reinforcing stability domains to more transient configurations of organisms and environment.

DOI

[4]
沈玉昌,苏时雨,尹泽生.中国地貌分类、区划与制图研究工作的回顾与展望[J].地理科学,1982,2(2):97-105.正 在自然地理环境中,地貌是基本要素之一,它与人类的生活、生产活动密切相关。因此,人类在很早以前就关心地貌,并加以分类、分区与制图。我国古代文献中虽然很早就有关于地貌的记载,但详细地叙述地貌并加以分类的著作,当首推二千多年前的《尔雅》一书,该书的“释地”、“释丘”、“释山”、“释水”等四篇,内容十分丰富详细。我国现今通用的平原、丘陵等名称均渊源于此。汉、唐以后,关于山川的记载更为详尽(如《汉书·地理志》)。宋代以来,方志之学蔚然兴

[ Shen Y C, Su S Y, Yin Z S.Retrospect and prospect of the research work on the classification, regionalization and mapping of the geomorphology[J]. Scientia Geographica Sinica, 1982,2(2):97-105. ]

[5]
姚永慧,周成虎,孙然好,等.基于多源数据的山地地貌数字解译[J].山地学报,2007,25(1):122-128.3S”技术与地学的结合,对传统地貌制图的编绘方法必然产生较大影响。遥感影像不仅作为地面调查的重要辅助手段,而且是一种可以根据需要缩放到所需比例尺的重要数据源;各种数据源的参考数据如地形图、植被图、地质图、土地资源图等均可以数字的方式在GIS环境中集成、叠加与运算,从而改变了地貌制图依赖于地形图进行多次综合概括的制图方式,使基于多源数据的遥感地貌信息定量解译成为可能。以青藏高原昌都幅遥感地貌解译为例,探讨不同类型的数据在山地地貌解译过程中的使用方式和对地貌判读的作用,并在此基础上总结基于多源数据的遥感地貌解译方法,为全国1:100万地貌遥感解译和地貌信息系统的建立奠定基础。

DOI

[ Yao Y H, Zhou C H, Sun R H, et al. Digital mapping of mountain landforms based on multiple source data[J]. Journal of Mountain Science, 2007,25(1):122-128. ]

[6]
闾国年,钱亚东,陈钟明.基于栅格数字高程模型提取特征地貌技术研究[J].地理学报,1998,53(6):562-570.本文对近年来基于栅格数字高程模型提取特征地貌技术进行了详细的研究,认为该技术的关键在于两个方面:一是如何定义地貌形态结构,二是提取算法的设计。本文提出了基于地貌学角度来定义地貌形态结构的方法,利用有限个数的形态要素的空间组合和对比分析来获取特征地貌,并可以对各种特征地貌形态进行符合物理意义的改进。

DOI

[ Lu G N, Qian Y D, Chen Z M.Automated extraction of the characteristics of topography from grid digital elevation data[J]. Acta Geographica Sinica, 1998,53(6):562-570. ]

[7]
刘爱利,汤国安.中国地貌基本形态DEM 的自动划分研究[J].地球信息科学,2006,8(4):8-14.我国1∶100万的数字高程模型,是在1∶5万及1∶10万基本 地形图上,高精度采集方里网交点高程所构建的1 km分辨率地面高程数字矩阵.本文利用该DEM数据及其所派生的多种地貌信息进行地貌形态类型自动划分的技术方法.实验提取地形起伏度、地表切割度、地表 粗糙度、高程变异系数、平均坡度、平均高程6个地形因子,并将各因子置于不同的信息层面中,通过主成分分析,ISODATA非监督分类法与 Bayesian最大似然监督分类法相结合,对中国地貌的基本形态进行了多维信息综合分类.研究结果表明:①我国1∶100万比例尺DEM在宏观地貌分类 方面具有重要的价值和应用潜力;②所提取的地形因子能宏观地反映我国地形的起伏特征,为地貌形态分类提供重要的依据;③采用ISODATA非监督分类法与 Bayesian最大似然监督分类法,能有效地实现我国地貌基本形态类型的定量化、自动化划分;④依据数据的统计特征进行分类,较合理地解决了类型模糊的 形态实体的归类问题.实验结果不仅揭示了此项技术在地貌形态分类中的巨大潜力,同时对于完善DEM数字地形分析的理论与方法也具有重要的意义.

DOI

[ Liu A L, Tang G A.DEM based auto-classification of Chinese landform[J]. Geo-information Science, 2006,8(4):8-14. ]

[8]
肖飞,张百平,凌峰,等.基于DEM 的地貌实体单元自动提取方法[J].地理研究,2008,27(2):459-466.lt;p>我国传统地貌基本形态类型分类强调地貌单元的完整性,界线划分沿地貌实体边界而非规则统计单元,目前尚缺乏地貌实体单元的有效自动提取方法。针对这一难点,本文提出一种基于DEM的地貌实体单元数字提取方法。利用坡度分级,并搜索相邻栅格单元、计算坡度级别内相互连通栅格的面积,建立坡度、面积阈值综合判别规则进行山地平原的自动划分;利用地形倒置、水文淹没分析,将山体划分的二维判别规则扩展到实际三维地形中,并结合地形结构线提取算法进行山体界线自动提取、确定山地地貌实体单元。结果表明,该方法符合我国传统地貌分类体系,能够较好实现山地/平原的自动划分和山体界线的数字提取。</p>

DOI

[ Xiao F, Zhang B P, Ling F, et al. DEM based auto-extraction od geomorphic units[J]. Geographical Research, 2008,27(2):459-466. ]

[9]
汤国安,杨玮莹,杨昕,等.对DEM地形定量因子挖掘中若干问题的探讨[J].测绘科学,2003,28(1):28-32.在地学研究中 ,地形结构信息的提取具有重要意义 ,而如何利用数字高程模型进行提取一直是地学工作者所面临的重要课题。在总结前人研究成果的基础上 ,从地形特征分析和水系特征分析两方面 ,比较了利用数字高程模型自动提取地形定量因子的基本原理、方法以及优缺点 ,并对其中存在的诸多理论与技术问题进行了系统的分析与探讨

DOI

[ Tang G A, Yang W Y, Yang X, et al. Some key points in terrain variables deriving from DEMs[J]. Science of Surveying and Mapping, 2003,28(1):28-32. ]

[10]
杨树文,谢飞,冯光胜,等.基于SOPT5图像的岩溶地貌单元自动提取方法[J].国土资源遥感,2012(2):56-60.通过对峰林、峰丛和岩溶洼地3者的地理特征和影像特征的研究,基 于遥感图像本底值提出了能有效反映目标特征的遥感指数——植被指数、土壤亮度指数、图像主成分变换第1主成分值及地形数据等,并构建了遥感指数的集成计算 法,建立了遥感自动提取模型.指数集成运算法能够有效地增大峰丛、峰林与其他地物之间的光谱差异,使这些岩溶地貌单元的灰度值高于其他地物,从而利于岩溶 地貌单元提取阈值的自动选取.基于构建的遥感自动提取模型先提取了峰丛、峰林信息,并在此基础上提取了岩溶洼地信息.经实验研究表明,该方法具有较高的提 取精度和效率.

DOI

[ Yang S W, Xie F, Feng G S, et al. Automatic extraction of Karst landscape elements based on SPOT5 image[J]. Remote Sensing for Land & Resources, 2012,2:56-60. ]

[11]
Dragut L, Blaschke T.Automated classification of landform elements using object-based image analysis[J]. Geomorphology, 2006,81:330-344.This paper presents an automated classification system of landform elements based on object-oriented image analysis. First, several data layers are produced from Digital Terrain Models (DTM): elevation, profile curvature, plan curvature and slope gradient. Second, relatively homogenous objects are delineated at several levels through image segmentation. These object primatives are classified as landform elements using a relative classification model, built both on the surface shape and on the altitudinal position of objects. So far, slope aspect was not used in classification. The classification has nine classes: peaks and toe slopes (defined by the altitudinal position or the degree of dominance), steep slopes and flat/gentle slopes (defined by slope gradients), shoulders and negative contacts (defined by profile curvatures), head slopes, side slopes and nose slopes (defined by plan curvatures). Classes are defined using flexible fuzzy membership functions. Results are visually analyzed by draping them over DTMs. Specific fuzzy classification options were used to obtain an assessment of output accuracy. Two implementations of the methodology are compared using (1) Romanian datasets and (2) Berchtesgaden National Park, Germany. The methodology has proven to be reproducible; readily adaptable for diverse landscapes and datasets; and useful in respect to providing additional information for geomorphological and landscape studies. A major advantage of this new methodology is its transferability, given that it uses only relative values and relative positions to neighboring objects. The methodology introduced in this paper can be used for almost any application where relationships between topographic features and other components of landscapes are to be assessed.

DOI

[12]
周旻曦,刘永学,李满春,等.多目标珊瑚岛礁地貌遥感信息提取方法——以西沙永乐环礁为例[J].地理研究,2015,34(4):677-690.南海珊瑚岛礁资源极为丰富,实时、快速、高效、准确地获取大范围珊瑚岛礁地貌遥感信息具有现实意义。研究提出了一种双尺度转化下的模型与数据混合驱动的岛礁地貌信息提取框架,并设计了珊瑚岛礁地貌分类体系及相应技术流程:首先采用自上而下模型驱动的GVF Snake模型进行宏观地理分带的粗分割,然后采用自下而上数据驱动的云影极值抑制下多阈值OTSU分类算法进行微观地貌类型的精细分类,最终利用区域生长算法提取离散分布的暗沙、暗滩等浅水地貌单元。针对西沙永乐环礁利用CBERS-02B数据进行实验,精度验证表明:珊瑚岛礁地貌遥感信息提取方法总体精度优于经典数据驱动的监督分类算法,且具有抗噪能力强、顾及空间拓扑关系、自动灵活等特点。

DOI

[ Zhou M X, Liu Y X, Li M C, et al. Geomorphologic information extraction for multi-objective coral islands from remotely sensed imagery: a case study for Yongle Atoll, South China Sea[J]. Geographical Research, 2015,34(4):677-690. ]

[13]
王岸,王国灿.构造地貌及其分析方法述评[J].地质科技情报,2005,24(4):7-12.构造地貌是指受构造内动力作用控制,通过内外地质动力的相互作用所奠定的能够反映一定构造特征的地貌形式。构造地貌学的研究内容为:地貌与构造的关系、构造地貌发生和发展过程以及构造地貌过程所揭示的地球内部构造动力过程;其分析方法可归纳为构造地貌格局分析法、构造地貌形态分析法、构造地貌相关沉积分析法和构造地貌年代分析法。构造地貌学从地形地貌的角度来分析构造过程,涉及不同圈层间的相互作用,响应了当前地球系统科学的研究思路,可以预见,构造地貌学将在圈层作用研究中发挥重要作用,同时朝着信息化、定量化的方向发展。

DOI

[ Wang A, Wang G C.Review on Morphotectonic and its analytical methods[J]. Geological Science and Technology Information, 2005,24(4):7-12. ]

[14]
张祖陆. 地质与地貌学[M].北京:科学出版社,2012:86-180.

[ Zhang Z L.Geology and geomorphology[M]. Beijing: Science Press, 2012:86-180. ]

[15]
杨景春,李有利.地貌学原理[M].北京:北京大学出版社,2012:175-184.

[ Yang J C, Li Y L.Principle of geomorphology[M]. Beijing: Peking University Press, 2012:175-184. ]

[16]
夏邦栋. 普通地质学[M].北京:地质出版社,1983:112.

[ Xia B D.Common geology[M]. Beijing: Geology Press, 1983:112. ]

[17]
彭丹青,李京,陈云浩,等.基于卫星遥感数据的地质断裂分形研究——以江西德兴为例[J].自然灾害学报,2008,17(6):119-123.利用分形理论分析了德兴地区断裂构造的特点.结果表明,不仅研究 区内的断裂带有自相似性,子区域也有相似现象.矿产区的分维值略高于整个研究区,铜矿区分维值高于金矿区,NW、NWW走向断裂带的分维值远高于SE走向 的.说明分维值越大,越有利于矿床的形成;分维值越大的断裂走向,对矿床的产出位置控制作用越明显;地质构造复杂程度越高,分维值也越大,线性展布也越复 杂.

DOI

[ Peng D Q, Li J, Chen Y H, et al. A fractal dimension study on geological fracture system based on remote sensing data: a case study of Dexing, Jiangxi[J]. Journal of Natural Disasters, 2008,17(6):119-123. ]

[18]
Dhont D, Chorowicz J, Yücrür T, et al. Polyphased block tectonics along the North Anatolian Fault in the Tosya basin area (Turkey)[J]. Tectonophysics, 1998,299:213-227.The Tosya basin is located in the bending segment of the North Anatolian Fault (NAF) in Turkey. We have obtained original observations on the neotectonics from SAR ERS images, Digital Elevation Model (DEM) and field structural analysis. Regional Neogene deformation is characterised by the occurrence of several basins that are superimposed in time and space. They result from differently oriented movements since 12 Ma, including southwestward motion along a fault subparallel to the NAF. We propose a model of polyphased tectonics related to the displacement of several individualised blocks. In the first stage (Tortonian), the North Tosya block has moved toward the N250° azimuth, parallel to the dextral N70°-striking segment of the NAF. As a consequence, a triple-junction-related compatibility basin was opened at the intersection with a N60° to N30°-striking fault. This pattern is similar to the Karliova corner where the NAF and the East Anatolian Fault meet. In the second stage (Early Pliocene–Middle Pleistocene), a segment of the former N70°-NAF was abandoned and the NAF propagated eastward to form a N90°-striking segment (N90°-NAF), cutting the former Tosya block and basin into two parts. The North Tosya block has moved again and this new geometry has permitted a South Tosya block to move parallel to the NAF but with a higher rate which has induced compression in the Tosya basin. In the third stage (Holocene), the South Tosya block moved toward N240°, obliquely to any of the NAF segments. This has resulted in the formation of two Holocene pull-apart type basins along the previous N60° to N30°-striking fault while extensional faults were formed in the South Tosya block. Estimated dextral displacement along the NAF is 5.9 to 8.5 km at this stage. This model of blocks moving in different directions, including Holocene local movements toward N240°, means that the NAF can be considered not to be a simple transform fault. Our model implies that the N90°-NAF was non-existent before the Early Pliocene.

DOI

[19]
Adiyaman Ö, Chorowicz J, Köse O.Relationships between volcanic patterns and neotectonics in eastern Anayolia from analysis of satellite images and DEM[J]. Journal of Volcanology and Geothermal Research,1998,85:17-32.

[20]
李德伦,王恩林.构造地质学[M].长春:吉林大学出版社,2001:238-240.

[ Li D L, Wang E L.Structural geology[M]. Changchun: Jilin University Press, 2001:238-240. ]

[21]
方世明,吴冲龙,刘刚,等.基于GIS的地质图图切剖面计算机辅助编绘[J].中国地质,2002(4):440-445.图切剖面是区域地质图的重要组成部分,用于反映地下的地层和构造 特征.传统的图切剖面是靠手工制作的,不但效率低下,而且不易保存和更新,作者研究开发了一种基于GIS的地质图图切剖面计算机辅助编绘系统.其编绘过程 主要包括以下几个步骤:(1)矢量化地质图的准备;(2)图切剖面的自动绘制;(3)人机交互修编与输出.本文以武汉市喻家山地质图的图切剖面编绘为例, 证明该系统具有高效率、高质量的特征,可以满足现有各种比例尺的地质图图切剖面的制作要求.

DOI

[ Fang S M, Wu C L, Liu G, et al. Computer-aided production of the cutting section of a geological map based on GIS[J]. Geology in China, 2002,4:440-445. ]

[22]
Nedas K A, Egenhofer M J.Spatial-scene similarity queries[J]. Transactions in GIS, 2008:12(6):661-681.Assessing spatial scenes for similarity is difficult from a cognitive and computational perspective. Solutions to spatial-scene similarity assessments are sensible only if corresponding elements in the compared scenes are identified correctly. This matching process becomes increasingly complex and error-prone for large spatial scenes as it is questionable how to choose one set of associations over another or how to account quantitatively for unmatched elements. We develop a comprehensive methodology for similarity queries over spatial scenes that incorporates cognitively motivated approaches about scene comparisons, together with explicit domain knowledge about spatial objects and their relations for the relaxation of spatial query constraints. Along with a sound graph-theoretical methodology, this approach provides the foundation for plausible reasoning about spatial-scene similarity queries.

DOI

[23]
Egenhofer M J.Query processing in spatial-query-by-sketch[J]. Journal of Visual Languages and Computing, 1997,8(4):403-424.Query processing in spatial-query-by-sketch EGENHOFER M. J. Journal of Visual Languages and Computing 8(4), 403-424, 1997

DOI

[24]
宋腾义,汪闽.多要素空间场景相似性匹配模型及应用[J].中国图象图形学报,2012,17(10):1274-1282.提出综合考虑空间对象的面积、属性及其相互间的拓扑、方位关系等 多要素的空间场景相似性匹配模型.首先利用“逐步转变模型”和属性关系图实现场景语义建模,而后构建场景匹配模型,利用所提出的相似度指标评价场景间的相 似性.该模型能够较为客观地刻画空间场景特别是地理场景的语义内涵,并且具有很好的场景匹配效果,在空间数据的智能查询检索上具有较好的应用前景.

DOI

[ Song T Y, Wang M.Multi-feature based spatial scene matching model and its application[J]. Journal of Image and Graphics, 2012,17(10):1274-1282. ]

[25]
Egenhofer M J, Al-Taha K K. Reasoning about gradual changes of topological relationships. Theories and methods of spatio-temporal reasoning in geographic space[C]. Proceedings of the International Conference GIS - from Space to Territory: Theories and Methods of Spatio-Temporal Reasoning in Geographic Space, 1992,639:196-219.

[26]
El-Kwae E A, Kabuka M R. A robust framework for content-based retrieval by spatial similarity in image databases[J]. ACM Transactions on Information Systems, 1999,17(2):174-198.A framework for retrieving images by spatial similarity (FRISS) in image databases is presented. In this framework, a robust retrieval by spatial similarity (RSS) algorithm is defined as one that incorporates both directional and topological spatial constraints, retrieves similar images, and recognizes images even after they undergo translation, scaling, rotation (both perfect and multiple), or any arbitrary combination of transformations. The FRISS framework is discussed and used as a base for comparing various existing RSS algorithms. Analysis shows that none of them satisfies all the FRISS specifications. An algorithm, SIMDTC, is then presented. SIMDTC introduces the concept of a rotation correction angle (RCA) to align objects in one image spatially closer to matching objects in another image for more accurate similarity assessment. Similarity between two images is a function of the number of common objects between them and the closeness of directional and topological spatial relationships between object pairs in both images. The SIMDTC retrieval is invariant under translation, scaling, and perfect rotation, and the algorithm is able to rank multiple rotation variants. The algorithm was tested using synthetic images and the TESSA image database. Analysis shows the robustness of the SIMDTC algorithm over current algorithms.

DOI

[27]
Aksoy S.Modeling of remote sensing image content using attributed relational graphs[J]. Lecture Notes in Computer Science, 2006,56:475-483.Automatic content modeling and retrieval in remote sensing image databases are important and challenging problems. Statistical pattern recognition and computer vision algorithms concentrate on feature-based analysis and representations in pixel or region levels whereas syntactic and structural techniques focus on modeling symbolic representations for interpreting scenes. We describe a hybrid hierarchical approach for image content modeling and retrieval. First, scenes are decomposed into regions using pixel-based classifiers and an iterative split-and-merge algorithm. Next, spatial relationships of regions are computed using boundary, distance and orientation information based on different region representations. Finally, scenes are modeled using attributed relational graphs that combine region class information and spatial arrangements. We demonstrate the effectiveness of this approach in query scenarios that cannot be expressed by traditional approaches but where the proposed models can capture both feature and spatial characteristics of scenes and can retrieve similar areas according to their high-level semantic content.

DOI

[28]
Eshera M A, Fu K S.A graph distance measure for image[J]. IEEE Transaction on Systems, Man, and Cybernetics, 1984,14(3):353-363.Attributed relational graphs (ARGs) have shown superior qualities when used for image representation and analysis in computer vision systems. A new, efficient approach for calculating a global distance measure between attributed relational graphs is proposed, and its applications in computer vision are discussed. The distance measure is calculated by a global optimization algorithm that is shown to be very efficient for this problem. The approach shows good results for practical size ARGs. The technique is also suitable for parallel processing implementation.

DOI

[29]
杜世宏,秦其明,王桥.空间关系及其应用[J].地学前缘 2006,13(3):69-80.

[ Du S H, Qin Q M, Wang Q.The spatial relations in GIS and applications[J]. Earth Science Frontiers, 2006,13(3):69-80. ]

[30]
艾廷华,郭仁忠.基于格式塔识别原则挖掘空间分布模式[J].测绘学报,2007,36(3):302-308.面向空间群目标的分布模式识别 是空间数据挖掘比较关注的问题。本研究基于空间认知原理与视觉识别格式塔完形原则并结合空间聚类方法对该问题进行研究,提出用于描述实体间差异的"视觉距 离"概念,其定义综合考虑视觉识别中的位置、方向、大小差异,通过Delaunay三角网计算几何构造建立该距离计算的模型。在实验基础上提出基于最小支 撑树MST的聚类方法,获得与视觉认知相一致的结果。研究试图表明一个观念,即通用性的数据处理模型在GIS实际应用时,需要根据GIS作为"空间认知" 科学的原理,作技术方法上的改进,需要考虑认知主体在感知、辨析、识别、推理不同思维过程中的认知心理原则。

DOI

[ Ai T H, Guo R Z.Polygon cluster pattern mining based on gestalt principle[J]. Acta Geodaetica et Cartographica Sinica, 2007,36(3):302-308. ]

[31]
Mark D M, Egenhofer M J.Modeling spatial relations between lines and regions: combining formal mathematical models and human subjects testing[J]. Cartography & Geographic Information Science, 1994,21(4):195-212.This paper describes the results of two human-subjects experiments to test how people think about spatial relations between lines and regions. The experiments focused on a formal model of topological spatial relations called the 9-intersection. For unbranched lines and simply connected regions, this model identifies 19 different spatial relations. Subjects were presented with two or three geometrically distinct drawings of each spatial relation (40 drawings in all), with the line and the region said to be a road and a park, respectively. In the first experiment, the task was to group the drawings so that the same phrase or sentence could be used to describe every situation in each group. A few subjects differentiated all 19 relations, but most identified nine to 13 groups. Although there was a great deal of variation across subjects in the groups that were identified, the results confirm that the relations grouped by the 9-intersection model are the ones most often grouped by the subjects. No consistent language-related differences were identified among 12 English-speaking subjects, 12 Chinese-speaking subjects, and four other subjects tested in their own native languages. A second experiment presented subjects with a short sentence describing a spatial relation between a road and a park, and the same 40 diagrams. Each subject was asked to rate the strength of their agreement or disagreement that the sentence described each relation. For each of the two different predicates tested—“the road crosses the park” and “the road goes into the park”—there was a great deal of consensus across the subjects. The results of these experiments suggest that the 9-intersection model forms a sound basis for characterizing line/region relations in the case of roads and parks, and that many spatial relations can be well-represented by particular subsets of the primitives differentiated by the 9-intersection.

DOI

[32]
Peuquet D J, Zhan C X.An algorithm to determine the directional relationship between arbitarily-shaped polygons in plane[J]. Pattern Recognition, 1987,20(87):65-74.The directional relationship between two polygons (e.g. left, above, beside, east, north) is an important spatial property and can also be used as a selection criterion for retrieving objects from a spatial database. If the database is large, this could help significantly in speeding search by reducing the size of the necessary search space. This paper builds upon past work to develop a model for determining the directional relationship in 2-D space between two simply-connected polygons of arbitrary shape, size and distance from each other. The model is based on visual interpretation and is data structure independent. The model is also stated in algorithmic terms, and is found to have a computational complexity of O ( n ), where n is the total number of vertices or cells used to represent the two polygons.

DOI

[33]
Mackaness W A, Edwards G.The importance of modelling pattern and structure in automated map generalization[C]. Joint Workshop on Multi-Scale Representations of Spatial Data, 2007:7-8.

[34]
Heinzle F, Anders K H.Characterising space via pattern recognition techniques: identifying patterns in road networks[J]. Generalisation of Geographic Information, 2007:233-253.

DOI

[35]
Robert C W, Zuell C, Landmann J, et al. Modality analysis: a semantic grammar for imputations of intentionnality in Texts[J]. Quality & Quantity, 2008,44(2):239-257.No abstract is available for this item.

DOI

[36]
Rao G, Agaewal C, Chaudhry S, et al. Natural language query processing using semantic gramma[J]. International Journal on Computer Science and Engineering, 2010,1(2):219-223.The field of natural language processing (NLP) has seen a dramatic shift in both research direction and methodology in the past several years. In the past, most work in computational linguistics tendedto focus on purely symbolic methods. Recently, more and more work is shifting toward hybrid methods that combine new empirical corpus-based methods, including the use of probabilistic and information theoretic techniques, with traditional symbolic methods.The main purpose of Natural Language Query Processing is for an English sentence to be interpreted by the computer and appropriate action taken. Asking questions to databases in natural language is a very convenient and easy method of data access, especially for casual users who do not understand complicated database query languages such as SQL. This paper proposes the architecture for translating English Query into SQL using Semantic Grammar.

[37]
Hopcroft J E.自动机理论、语言和计算机导论[M].北京:机械工业出版社,2012:120-122.

[ Hopcroft J E.Introdution to automata theory, languages, and computation third edition[M]. Beijing: China Machine Press, 2012:120-122. ]

[38]
程维明,周成虎.多尺度数字地貌等级分类方法[J].地理科学进展,2014,33(1):23-33.参考已出版的全国各级各类比例尺的地貌类型图的分类方案及图例表达,探讨了中国1∶100万数字地貌的等级分类方法,采用基于形态、成因、物质和年龄等地貌要素,综合反映地貌特征的等级分类指标和分类体系,初步构建了中国多个国家基本比例尺(即1∶400万、1∶100万、1∶50万、1∶25万、1∶5万)数字地貌等级分类方法,发展了由连续分布的多边形图斑反映形态成因类型,以及由离散的点、线和面图斑共同反映形态结构类型的数字地貌类型数据组织方式,构建了多尺度数字地貌类型的编码方法.该研究可为发展多尺度地貌类型图的编制提供方法基础,也可为当前正在进行的地理国情监测工程的大比例尺地貌类型信息普查提供分类规范和技术支持.

DOI

[ Cheng W M, Zhou C H.Methodology on hierarchical classification of multi-scale digital geomorphology. progress in geography[J]. Progress in Geography, 2014,33(1):23-33. ]

[39]
张根寿,林爱文,王新生,等.庐山地理调查[M].武汉:武汉大学出版社,2004:6-7.

[ Zhang G S, Lin A W, Wang X S, et al.Geographical investigation in Lushan[M]. Wuhan: Wuhan University Press, 2004:6-7. ]

[40]
张兆干,王腊春.庐山地区地理学野外实习指南[M].北京:科学出版社,2011:27-28.

[ Zhang Z G, Wang L C.Guideline of geographical fieldwork in Lushan area[M]. Beijing: Science Press, 2011:27-28. ]

[41]
中华人民共和国地质矿产部. DZ/T 0197-1997, 中华人民共和国地质矿产行业标准[S].北京:中国标准出版社,1998.

[ DZ/T 0197-1997Z/T 0197-1997. Standard of the People's Republic of China on the Geological and Mineral Industry[S]. Beijing: China Standard Press, 1998. ]

[42]
海会幅地质图说明书[R].会幅地质图说明书[R].江西地质矿产勘查开发局,1994:54-55.

[ Geological introduction of Haihui area[R].eological introduction of Haihui area[R]. Nanchang: Jiangxi Bureau of Geology and Mineral Resources, 1994:54-55. ]

Outlines

/