Designing a perception based anti-aliasing filter for enhancement of down-sampled images

Abstract

In this paper, the problem of aliasing due to pixel based image down-sampling in CMYK color space is addressed. Such a problem exists when a high-resolution image or video is to be mapped to low-resolution. Signal processing theory tells us that optimal decimation requires low-pass filtering with a suitable cutoff frequency, followed by down-sampling which remove many useful image details blurring. Instead of operating in the entire image, the proposed method finds the edge maps and then applies anti-aliasing filters only on the edge map regions excluding the horizontal and vertical edges. The algorithm shows a significant reduction of the aliasing artifacts, commonly known as "jaggies". Perceptual relative color dominance which is calculated from psycho visual experiments is included in the anti-aliasing part to improve the performance of the algorithm. The number of color quantization levels is varied separation by separation for each color channel and psycho visual survey is conducted to find the perceptual color dominance. © 2013 Springer.