Title
Image reconstruction using Loop subdivision
Document Type
Conference Proceeding
Publication Date
12-1-2010
Abstract
It is an ultimate objective to reconstruct an image with high quality using a compact representation, which is the basic step in image-manipulation fields. We propose an effective vectorization based approach to reconstruct an image using a triangular mesh associated with Loop subdivision scheme in the present paper. With an initial control mesh obtained by simplifying a dense mesh from a quality-preserved triangulation, we produce the final optimal control mesh by optimizing a mesh over topologies and colors to approximate the given image. The main advantages of the approach include: (1) the reconstruction of an image is not restricted to be aligned with image coordinate axes; (2) a high order continuous function is defined over a triangle instead of a bilinear interpolation; (3) it is a compact and vector-based representation easy to edit and transmit. Experimental results are presented to confirm the effectiveness of the method. Comparisons with the bi-cubic spline and the mesh simplification methods demonstrate the merits of our method in reconstruction quality and representation size.
Publication Source (Journal or Book title)
APSIPA ASC 2010 - Asia-Pacific Signal and Information Processing Association Annual Summit and Conference
First Page
974
Last Page
982
Recommended Citation
Zhou, H., Zheng, J., & Li, X. (2010). Image reconstruction using Loop subdivision. APSIPA ASC 2010 - Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, 974-982. Retrieved from https://repository.lsu.edu/eecs_pubs/791