Novel ESNIU upsampling method for lossy image compression

Abstract

Lossy image compression is an attractive field for researchers. Image interpolation and demosaicking concepts can induce the higher compression ratio of lossy image compression. This paper proposes an image compression method which comprises of the Novel scheme for image-upsampling namely ‘Error Signal Constrained Neighbor based image upsampling (ESNIU)’. The existing demosaicking method namely ‘Multiscale Gradients based color filter array Demosaicking (MGD) is employed to achieve the demosaicking. The combination of these two methods emerge the high-grade proposed method for lossy image compression and the new method is abbreviated as ESNIU-MGD. In this paper, the main contribution ESNIU upsampling is designed to effectively handle the Green-Error signals for upsampling with the aid of neighbour pixels. The experimental results show that the proposed ESNIU-MGD method makes significant positive enhancement in both cases of quality and compression ratio. The PSNR and CR of the ESNIU-MGD proves the promising enhancements than the existing lossy image compression methods.