Radviz-based visual analysis of multispectral images

The technique of multispectral imaging has gained growing attention in recent years. This is mostly due to the potential to provide rich information that can be used to improve material classification or product quality assessment. Due to the complex nature of the corresponding datasets, processing tools are needed to gain insights into the data. In this paper, we propose an analysis tool to assist a user with the interactive evaluation of multispectral images. The aim of the approach is to provide easier access to the wealth of information and to obtain a segmented multispectral image. Multivariate radial- and image-based visualizations, are meaningfully combined by linked views to find a proper segmentation in a semi-automatic way. The linked views are complemented by a novel evaluation view that allows the evaluation and the refinement of the segmentation, if necessary. We show usage examples of our proposed processing approach for multispectral scene data.