Quality Enhancement of Direct Volume Rendered Images

Abstract

In this paper, we propose a new method for enhancing the quality of direct volume rendered images. Unlike the typical image enhancement techniques which perform transformations in the image domain, we take the volume data into account and enhance the presentation of the volume in the rendered image by adjusting the rendering parameters. Our objective is not only to deliver a pleasing image with better color contrast or enhanced features, but also generate a faithful image with the information in the volume presented in the image. An image quality measurement is proposed to quantitatively evaluate image quality based on the information obtained from the image as well as the volumetric data. The parameter adjustment process is driven by the evaluation result using a genetic algorithm. More informative and comprehensible results are therefore delivered, compared with the typical image-based approaches.