DocumentCode :
3198261
Title :
Adopt adaptive B-spline to embellish contours in image segmentation
Author :
Lee, Bin ; Zhuang, Tian-ge
Author_Institution :
Dept. of Biomed. Eng., Shanghai Jiaotong Univ., China
fYear :
2001
fDate :
2001
Firstpage :
216
Lastpage :
221
Abstract :
Image segmentation is an important process in computer vision. Most of the currently popular methods tracking the object edge depend on the pixel gray value. Because of the noise in images, the pixel resolving and the precision of the detection method itself, the detected contours are far from smooth and with many tiny zigzags, which will affect the subsequent processing, such as image recognition, image feature detection, etc. This paper puts forwards a novel method to get a smooth contour in the image segmentation process. Firstly, adopt a dynamic programming algorithm to process the object image and to get a global optimal boundary, then smooth and fit the boundary by an adaptive B-spline, which adjusts the control points depending on the contour curvature. Practical numerical experimental results show that this method has stronger edge detection ability and has produced smoother contour curves than other methods and without losing the edge feature at the same time
Keywords :
dynamic programming; edge detection; image segmentation; splines (mathematics); adaptive B-spline; computer vision; contour curves; dynamic programming algorithm; edge detection; experimental results; global optimal boundary; image feature detection; image recognition; image segmentation; object edge tracking; pixel gray value; smooth contour; smoothing; Computer vision; Dynamic programming; Heuristic algorithms; Image edge detection; Image recognition; Image resolution; Image segmentation; Pixel; Programmable control; Spline;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Medical Imaging and Augmented Reality, 2001. Proceedings. International Workshop on
Conference_Location :
Shatin, Hong Kong
Print_ISBN :
0-7695-1113-9
Type :
conf
DOI :
10.1109/MIAR.2001.930289
Filename :
930289
Link To Document :
بازگشت