Eptascape MPEG-7 Encoder

Developed by a professional team of scientists, the Eptascape MPEG-7 encoder provides a fully automated way to generate a real-time description of the video in terms of objects and their features.

Typical output is a set of low-level "Visual Descriptor Tools" for each individual video frame, encoded in XML and according to the ISO/IEC 15938-3 standard. Based on a five-stage pipeline, the encoder produces other intermediate outputs to be used in specialized applications.

Designed to be used either as a component of an embedded system or as SDK for hosted applications, the Eptascape MPEG-7 encoder is used in a variety of video analytics applications where automatic event detection in real-time is essential.

The MPEG-7 encoder is available for the Texas Instruments TMS320DM642 and Da Vinci processors with a parsing library for the Linux and MS-Windows operating systems. On a TI DM642 operating at 600 MHz, the encoder can process 15-30 frames per second of a CIF (352 x 240) video feed and has a footprint of about 200 KBytes.

Read more about MPEG-7 ...

See the encoder in action with behavior detection ...