Affiliation:
1. Northeast Petroleum University
2. Dalian University of Technology
Abstract
Image interpolation is used widely in the computer vision. Holding edge information is main problem in the image interpolation. By using bilinear and bicubic B-spline interpolation methods, a novel image interpolation approach was proposed in this paper. Firstly, inverse distance weighted average method was used to reduce image’s noise. Secondly, edge detection operator was used to extract image's edges information. It can help us to select different interpolation methods in the image interpolation process. Finally, we selected bilinear interpolation approach at non-edge regions, and bicubic B-spline interpolation method was used near edges regions. Further more, control vertexes were computed from pixels with calculation formula which has been simplified in the B-spline interpolation process. Experiments showed the interpolated image by the proposed method had good vision results for it could hold image's edge information effectively.
Publisher
Trans Tech Publications, Ltd.
Reference7 articles.
1. K. Jensen, D. Anastassion. Subpixel Edge Localization and the Interpolation of Still Images. IEEE Trans. on Image Processing, 1995, 4, (3), pp.285-295.
2. B. Su, D. Liu. Computational Geometry, Shanghai, Shanghai Scientific and Technical Publishers, (1980).
3. R. Zhou, Z. Su, Y. Xi, S. Shao. Curve and Surface in CAGD. Jilin University Press, (1993).
4. Y. Yang, Z. Su, J. Hu. A New Edge-holding Algorithm of Image Interpolation, Journal of Image and Graphics, 2005, 10, (10), pp.1248-1251.
5. C.X. Durand, D. Faguy. Rational zoom of bitmaps using B-spline interpolation in computerized 2-D animation, Computer Graphics Forum, 1990, 9, (1), pp.27-37.