Vectorization in graphics recognition: to thin or not to thin
Authors:
- K. TombreKTK. Tombre
- S. TabboneSTS. Tabbone
- 11 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 91-96
Abstract
Vectorization, i.e. raster-to-vector conversion, is a cen- tral part of graphics recognition problems. In this paper, we discuss the pros and the cons of basing one's vectoriza- tion process on skeletonization. While distance skeletons have proven to be robust and precise, they tend to distort the results at line extremities and junctions. In these cases, contour-matching approaches yield better results, but they have their own specific problems. A perspective is probably to combine the best of both methods.Keywords
This publication has 19 references indexed in Scilit:
- Raster-to-vector conversion by line fitting based on contours and skeletons2002
- Feature analysis using line sweep thinning algorithmIEEE Transactions on Pattern Analysis and Machine Intelligence, 1999
- Skeletonization of binary images with nonuniform width via block decomposition and contour vector matchingPattern Recognition, 1998
- Adaptive Vectorization of Line Drawing ImagesComputer Vision and Image Understanding, 1997
- Skeleton generation of engineering drawings via contour matchingPattern Recognition, 1994
- Generating skeletons and centerlines from the distance transformCVGIP: Graphical Models and Image Processing, 1992
- Thinning methodologies-a comprehensive surveyIEEE Transactions on Pattern Analysis and Machine Intelligence, 1992
- High Quality Vectorization Based on a Generic Object Model1992
- A robust algorithm for text string separation from mixed text/graphics imagesIEEE Transactions on Pattern Analysis and Machine Intelligence, 1988
- A Computational Approach to Edge DetectionIEEE Transactions on Pattern Analysis and Machine Intelligence, 1986