Studying Effects of Color Quantization an Image Zooming

Abstract

Color image quantization, a form of lossy image compression, is the process of reducing the number of unique colors in an image without appreciable reduction in quality. Originally color quantization was developed for satisfying hardware constraints of the display devices that allowed only a limited number of colors to be displayed simultaneously. In this work primarily motivated by an attempt to study how the number of colors and the physical size of an image (in pixels) are related and we try to use a new algorithm, which reduces colors (apply quantization) when we zoom out the image and increases colors (restore lost colors) when we zoom in the image (dynamic color quantization). Our experiments show the relationship between dynamical quantization and image quality using different approaches.