This thesis investigates how content-based indexing and retrieval systems can be used to analyse digital video. We focus particularly on the challenge of applying colour-analysis methods to large amounts of heterogeneous television broadcast video. Content-based systems are those which attempt to automatically analyse image or video documents by identifying and indexing certain features present in the documents. These features may include colour and texture, shape, and spatial locations. Digital video has become hugely important through the widespread use of the Internet and the increasing number of digital content providers supplying the commercial and domestic markets. The challenge facing the indexing of digital video information in order to support browsing and retrieval by users, is to design systems that can accurately and automatically process large amounts of heterogeneous video. The basic segmentation of video material into shots and scenes is the basic operation in the analysis of video content. Although many published methods of detecting shot boundaries exist, it is d ifficult to compare and contrast the available techniques. This is due to several reasons. Firstly, full system implementation details are not always published and this can make recreation of the systems difficult. Secondly, most systems are evaluated on small, homogeneous sequences of video. These results give little indication how such systems would perform on a broader range of video content types, or indeed how differing content types can affect system performance. As part of an ongoing video indexing and browsing project, our research has focused on the application of different methods of video segmentation to a large and diverse digital video collection. A particular focus is to examine how different segmentation methods perform on different video content types. With this information, it is hoped to develop a system capable of accurately segmenting a wide range of broadcast video. Oilier areas addressed in this thesis include an investigation of evaluation methods for digital video indexing systems, and the use of adaptive thresholds for segmentation of video into shots and scenes.