The present invention relates to the detection of a predetermined sequence in a digital bit-stream, and more particularly a method and apparatus for the fast and efficient detection of a start code sequence.
As with many packet based bit-streams, packets are identified through the use of a start code. The start code is a unique sequence which occurs only to indicate the start of a packet, and can never occur in the data portion of a bit-stream. Identifying the start of packets is crucial in the processing of packetised bit-streams.
In the field of digital broadcasting, a common format of digital video compression is that of the Moving Picture Expert Group (MPEG). MPEG uses a packetised bit-stream and packets are preceded by a start code to enable individual packets to be identified. In any real-time processing of MPEG bit-streams, it is vital to be able to identify the MPEG start codes as quickly and efficiently as possible. Performing this in hardware is a relatively straightforward operation. Detecting start codes using software is also straightforward in a non real-time situation. However, where an end-to-end real-time software solution is required to process MPEG data it may not be possible or desirable to use a hardware-based solution.
The present invention overcomes the problems of the prior art and provides a method and apparatus for the fast and efficient detection of the MPEG start code sequence.