Area efficient lagrange image up scaling architecture for multimedia applications

Abstract

Interpolation is the process of constructing new pixels from the known pixels in an image. This work presents an area efficient up-scaling architecture of Lagrange interpolation for digital image. Lagrange interpolation is widely used to get high visual quality. It is a technique to estimate the unknown pixel values from the given pixels. It provides high estimation accuracy and fewer additions and shift operations are needed to calculate the weighting coefficients. To enhance the visual quality, clamp filter is implemented with the interpolation technique. It is used to preserve the edge information. The Hardware architecture of Bi-cubic, Lagrange interpolation algorithm and clamp filter based bi-cubic, Lagrange interpolation algorithm for various kernels such as 3×3 kernel, cross model, 2×3 model and T-model was simulated in MATLAB Simulink and Xilinx ISE generator. Among this various kernels of clamp filter, T model clamp filter with Lagrange interpolation achieved 10.52% reduction in LUTs compared to Bi-cubic interpolation algorithm.