Data Fusion-based Spatio-temporal Adaptive Interpolation for Low-resolution Video

This paper presents an adaptive regularized image interpolation algorithm from blurred and noisy low resolution image sequence, which is developed by a general framework based on data fusion. This framework can preserve edge information and suppress noise, effectively. To obtain a higher resolution image, both discontinuities across the edge orientation and smoothness along the edge orientation should be considered in a high resolution image interpolation algorithm. Also, spatio-temporal adaptive three dimensional Laplacian smoothness constraints are used to suppress noise and preserve edge profiles. This multichannel image interpolation algorithm is to construct an adaptive fusion algorithm based on regularized image interpolation. By combining spatially adaptive constraints, directional high frequency components are preserved with efficiently suppressed noise. Experimental results show that the proposed algorithm has the advantage of preserving directional high frequency components and suppressing undesirable artifacts such as noise.