This paper describes a solution for the organization and management of multimedia collections in digital libraries. Video U-DL-A (VUDLA) is an extension to a digital library that allows for storage, indexing and annotation of multimedia documents. It functions in such way that text- and image-based queries can be issued in order to retrieve specific scenes from digital video collections. Technologies such as image and speech processing, video streaming, multimedia databases, information retrieval and graphical user interfaces are integrated to produce a novel multimedia, multimodal environment which re-evaluates text as an important medium for knowledge transmission. We have developed a fully operational testbed to explore multimedia data properties and organization possibilities as well as a wide range of practical applications.