Convera has applied two decades of experience in intelligent, high performance search and retrieval technology to develop the world’s most sophisticated and easy-to-use video asset management system. The result is Screening Room, which gives today’s video producers a fast, accurate, fully integrated, modular system to browse, search and preview all of their video source material (either analog or digital) directly from their desktops and in a fraction of the time it takes to locate material by traditional methods.
Video producers can effortlessly search vast tape archives for supplemental footage; automatically capture video; browse storyboards; catalog content using annotations, closed captioned text, voice sound tracks, and metadata; search for precise video clips using text and image clues; create rough cuts and edit decision lists(EDLs) for further production; and publish those video assets to the Web for streaming.
Screening Room is a cost-effective way to turn video assets into useful, profitable corporate resources.
Visual RetrievalWare SDK is a software developer's kit for building image indexing and pattern recognition applications. The VRW SDK is a multi-platform, thread-safe, software developer's kit for image processing; feature extraction, indexing, and retrieval; associative data mapping; and video analysis. It is intended for use in a wide spectrum of image indexing and management applications.
The VRW SDK is written in C, and includes language bindings for Tcl/Tk. The SDK is organized into several semi-interdependent C APIs. The IPS C API supports reading and writing most popular image file formats, as well as cutting, pasting, and otherwise processing images. The FEX C API extracts different types of features from still images to produce feature vectors. The FVS C API and FVI C API are used to create feature vector indexes, which allow a single feature vector to be compared against large numbers of previously indexed feature vectors very quickly. The MDB C API provides simple associative database functionality that can map arbitrary binary keys into whatever data you wish to associate with them, which is useful for example in mapping FVI object identifiers into filenames and vice versa. The VAE C API implements a video analysis engine that extracts information from video and can, for example, be used to automatically generate storyboards.
In addition the SDK includes several C APIs implementing data structures shared by the functional C APIs described above. The CTX C API provides thread-specific context, including pluggable system calls and error logging. The FV C API provides feature vectors. And the PKG C API provides machine-independent data marshalling for all C API data structures.
The VRW SDK contains the core image processing, indexing, and feature extraction algorithms developed and used by Convera over the past decade. Convera has used these algorithms and their progenitors to demonstrate solutions to many image-related problems, including: video clip matching, color/shape/texture image indexing and retrieval, handwritten kanji character recognition, forms recognition, face recognition, fingerprint recognition, and more. The VRW SDK is also directly applicable to other problems involving the comparison of large numbers of images, including: surveillance, classification, image database, quality control, and template matching.
|
