A Gradient-Based Adaptive Interpolation Filter for Multiple View Synthesis

Authors:
Ping Yang;Xin Tong;Xiaozhen Zheng;Jianhua Zheng;Yun He
Affiliations:
State Key Laboratory on Microwave and Digital Communications, Tsinghua National Laboratory for Information Science and Technology, Dept. of EE, Tsinghua University, Beijing, China 100084;State Key Laboratory on Microwave and Digital Communications, Tsinghua National Laboratory for Information Science and Technology, Dept. of EE, Tsinghua University, Beijing, China 100084;State Key Laboratory on Microwave and Digital Communications, Tsinghua National Laboratory for Information Science and Technology, Dept. of EE, Tsinghua University, Beijing, China 100084;State Key Laboratory on Microwave and Digital Communications, Tsinghua National Laboratory for Information Science and Technology, Dept. of EE, Tsinghua University, Beijing, China 100084;State Key Laboratory on Microwave and Digital Communications, Tsinghua National Laboratory for Information Science and Technology, Dept. of EE, Tsinghua University, Beijing, China 100084
Venue:
PCM '09 Proceedings of the 10th Pacific Rim Conference on Multimedia: Advances in Multimedia Information Processing
Year:
2009

Citing 6
Cited 0

Bilateral Filtering for Gray and Color Images

ICCV '98 Proceedings of the Sixth International Conference on Computer Vision
A fast approximation of the bilateral filter using a signal processing approach

ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part IV
Warped distance for space-variant linear image interpolation

IEEE Transactions on Image Processing
Edge-directed prediction for lossless compression of natural images

IEEE Transactions on Image Processing
New edge-directed interpolation

IEEE Transactions on Image Processing
A Fast Image Super-Resolution Algorithm Using an Adaptive Wiener Filter

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

A gradient-based adaptive interpolation method is proposed in this paper to solve the over-blurring problem in conventional multiple view synthesis (MVS) filters. To improve the visual quality of final synthetic pictures, a good interpolation filter is required in multiple view synthesis steps. Traditional space-invariant filters, such as bi-linear or bi-cubic filter, take the advantage of complexity, but they also lead to a decrease of the subjective quality. In contrast, directional filters usually exploit the directional information, especially in edge area, to deal with the over-blurring problem. This paper proposes a fast directional interpolation method for scaling up the resolution or filling up the losing pixels of a picture. The gradient map of an input picture is calculated in first. Using the gradient of each input pixel, the interpolation coefficients computed from a Gaussian kernel is refined, leading to a directional filter which takes an adaptation with gradient direction. This method is with a low complexity because it is a non-iterative method, and experiment results show that the visual quality of interpolated picture is improved.