DocumentCode :
3109178
Title :
Bezier Curve-Based Character Descriptor Considering Shape Information
Author :
Sohel, Ferdous Ahmed ; Karmakar, Gour C. ; Dooley, Laurence S.
Author_Institution :
Monash Univ., Churchill
fYear :
2007
fDate :
11-13 July 2007
Firstpage :
212
Lastpage :
216
Abstract :
While a Bezier curve (BC)-based character descriptor calculates the control points (CPs), it does not consider the shape characteristics for a prescribed distortion. This paper presents a novel Bezier Curve-based character descriptor considering shape information (BCDSI) which automatically divides a shape into segments based on the curvature and cornerity properties. Each segment is then defined by either a BC or straight line, depending on the distortion criteria in terms of a set of CPs. Quantitative and qualitative performances of the algorithm have been rigorously analysed upon both English and Bangla alphabetic characters, with the results confirming that the new BCDSI algorithm exhibits superior performance over other currently available shape descriptor algorithms employing the BC.
Keywords :
character recognition; curve fitting; edge detection; feature extraction; image coding; image segmentation; BC-based character descriptor; Bezier curves; CP encoding; control points; cornerity properties; curvature properties; shape information; shape segmentation; Algorithm design and analysis; Australia; Automatic control; Communication system control; Encoding; Information technology; Mobile communication; Multimedia communication; Performance analysis; Shape control; Bezier curves; control points.; corner points; curvature; shape descriptor;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer and Information Science, 2007. ICIS 2007. 6th IEEE/ACIS International Conference on
Conference_Location :
Melbourne, Qld.
Print_ISBN :
0-7695-2841-4
Type :
conf
DOI :
10.1109/ICIS.2007.69
Filename :
4276383
Link To Document :
بازگشت