CISRG Projects

Terrain Visualisation

• Exhibition on Art in Scientific Visualization of Terrain Data
  by Mahes Visvalingam, initially mounted at the Royal Institution on 5 November 1999 and 17 March 2000.

The following projects have been published since then.

• Sketch-based evaluation of Line Filtering Algorithms
  Paper by Mahes Visvalingam presented at GIScience 2000, Savannah, Georgia, USA (28 - 31 Oct, 2000) - published by Visvalingam and Dowson (2001) in The Visual Computer.
  
• Formulated Silhouettes for Sketching Terrain
  Paper by John Whelan and Mahes Visvalingam.
  • Terrain Sketching for Visual Insights Inc. - the Gettysburg Battlefield by Mahes Visvalingam and John Whelan
    
• Silhouettes in 3D
  • Evaluation of some published algorithms :
    • Cass Everitt's One Pass algorithm - the cubemap_toon demo - evaluated by Ruzi Che Mat and Mahes Visvalingam
    • Rossignac's Two Pass Algorithm for Wireframe Silhouettes and Raskar and Cohen's Polygon Fattening - the Sil2 demo - evaluated by Ruzi Che Mat, James Ward and Mahes Visvalingam
  • The Semiology of Silhouettes and related terms (explains the origins of discussion thread and lists the participants and contributions).
    - inaugurated by Mahes Visvalingam (CISRG) and Ramesh Raskar (MERL).  
      Please email Mahes Visvalingam with comments and/or request to be added to the list of participants.

Theoretical Cartography (Under construction)
This sub-web contains Mahes Visvalingam's contributions to the work of the British Cartographic Society and the International Cartographic Associations' Commission on Theoretical Cartography.  The papers are philosophical and deal with definitions, map-induced semiosis, cartosemiotics, etc.  The pages contain links to other sites.

• Contents Page (Last revised Mar 2003)
  • Definitions (Last revised Jan 2003)
  • Scope of Digital Cartography (Uploaded June 2003)
  • Cartosemiotics (Last revised April 2003)
  • Visualisation, ViSC and Scientific Insight (Uploaded June 2003)

Census Analysis and the Role of Maps in Social Indicator Research

• Data Rich - Information Poor
  Lecture by Mahes Visvalingam on Friday October 20, 2000  at IAP2K, the annual symposium of the UK Institution of Analysts & Programmers, which was   held in association with the British Computer Society at Trinity House, London.  The talk focused on the problems involved in defining statistical indicators of local conditions. 

Line Generalisation

• Phil Wade's Fortran code for The Douglas-Peucker Algorithm in CISRG DP 4.
• Evaluation of the Douglas-Peucker Algorithm by Visvalingam and Whyatt (1990) CISRG DP 6.
• Cartographic Algorithms: problems of implementation and evaluation and the impact of digitising errors in CISRG DP 8
• Visvalingam's algorithm for line generalisation in CISRG DP 10
• Other Comparisons of Visvalingam's algorithm with Douglas-Peucker and Bendsimplify algorithms
  - Visvalingam and Williamson (1995) in Cartography and GIS - CISRG DP13
  - Visvalingam and Herbert (1999) in Cartography and GIS - original pre-published CISRG DP 16 being scanned.
• Line Generalisation Algorithms and Terrain Visualisation. (see above)
• Deconstructing Fractals
  - Exercises undertaken by members of the Design Group of the British Cartographic Society Visvalingam (1999) in Cartographic J
  - also paper by Visvalingam and Brown (1999)
  - implications for cartographic education in CISRG DP 17
• The need to segment line features into semantic areal entities for map generalisation - presented by Visvalingam at ICA 1999; page includes links to:
  - a rather rambling draft paper seeking to explain my position and approach
  - powerpoint slides used at ICA
  - illustrations were based on Herbert's MSc project on line segmentation.
  This paper linked my interests in generalisation with our research in spatial data modelling.

Spatial Data Modelling and Object Recognition (Under construction - last revised September 2005)

This sub-web was created in response to a number of recent email enquiries on CISRG research, dating from 1985 to 1994, on spatial data models for GIS.   We undertook a series of investigations  in collaboration with the Ordnance Survey into map data models and algorithms for validating link-and-node structured vectors and for automatic extraction of implied areal topographic objects. The recently launched Ordnance Survey (OS) MasterMap^TM uses unique Topographic Identifiers (TOID) and National Grid references for management and cross-linking of disparate geo-referenced databases based on primitive regions as suggested by the CISRG in the mid 1980s.

Relevant CISRG references are:

1. Areal Units and the Linking of Data : some conceptual issues
   by Mahes Visvalingam, Chapter 2 in Worrall L (ed) Spatial Analysis and Spatial Policy Using Geographic Information Systems (Belhaven Press, London), 1991, 12 - 37.  This chapter, posted here with permission from the editor, is a review of the issues involved in the design of basic spatial units (BSUs).
2. CISRG Discussion Paper 1 entitled From Line Geometry to Area Topology provides a detailed account of our research on spatial data modelling and algorithms for casting data.  These results were published in The Computer Journal.  Readers may wish to note the critique of our approach by Laurini and Thompson (1992).
3. CISRG Discussion Paper 11 entitled Area Topology for Road Extraction and Topographic Data Validation demonstrates an application of the Spatial Data Model described in CISRG DP 1 for extracting and validating incompletely defined topographic objects from link-and-node structured map data and thereby detecting topological errors in the source data and deficiencies in the map model specification respectively.
4. CISRG Discussion Paper 12 entitled Global dimensions for the recognition of prototypical urban roads in large scale vector topographic maps noted that three global dimensions, including two novel ones proposed by the CISRG, are useful for recognising prototypical roads from the geometric information alone in link-and-node structured map data. Roads extracted through use of semantic codes (see CISRG DP11) can then be checked against the results of recognition to provide yet another means for validating topographic data.

Promotion of Independent Research (last revised March 2003 )
The research undertaken by Mahes Visvalingam has been independent and largely unfunded.  As such, she is keen to encourage the efforts of other independent researchers through this sub-web (last updated on Mar 2003).

Cartographic Information Systems Research Group, University of Hull