De-Noising the Image Using B-Spline Interpolation

Abstract

This paper reconstructs the perfect resolution for image using B-spine through four different steps to decompose the image and the sub-bands of the image incorporating the threshold function. B-spline polynomial interpolation for sub-band image along with three different convolutions like convolution of n - 1 and n, convolution of n and n + 1 and convolution of n + 1 and n - 1 calculated along with B-spline to compute the noise. To determine this type of recursive noise and enhance the noisy image as noise free image we introduced the separate B-spline bmn kernel, which is obtained by sampling the stretched B-spline degree n and factor m. Further, Cardinal spline is associated with the normal B-spline interpolation subject to the matrix framework and establishing a system of tri diagonal and band diagonal system of equations on band-limited functions. Finally numerical and optimal basis of B-spline of degree n = 3 is exposed.