A Novel Method for Image Compression Using Spectrum

TLDR: This work proposes a novel spectrum based image compression technique, which first blur the image with Point Spread Function determined using spectrum of given image, and then performs deconvolution using the known blur PSF to recover the image.

Abstract: Compression is an important aspect of image processing with respect to transmission and storage. It is an established field of research with very little scope for improvement through customary coding based compression techniques. Consequently, non-customary ways of compression have become an important area for future research. Any information that can be restored can be compressed. Using this principle, we propose a novel spectrum based image compression technique. We first blur the image with Point Spread Function (PSF) determined using spectrum of given image. Blurring increases the DC component in the image which in turns gets further compressed by DCT based JPEG compression compared to original image. To recover the image, we perform deconvolution using the known blur PSF.