A method for demultiplexing digital video data which originates from multiple sources in which the multiplexed images are not indexed nor is there identification information provided to differentiate sources. The sources are generally cameras which may be stationary cameras or moving cameras that rotate as are commonly used in the surveillance industry. A first set of digital video data representative of a first image is retrieved from a memory source or from a video tape. The first set of digital video data is stored to a memory location associated with a first source. The first set of video data is also identified as representative video data of the first source. A second set of digital video data representative of a current image is then retrieved. A difference ratio is calculated using the representative digital video data and the current set of digital video data. If the difference ratio is above a threshold, a display query requiring response for classifying the current digital video data as belonging to an identified source or originating from a new source is presented to a user of the system. If a response is received indicating that the current set of digital video data is associated with a new source, the current set of digital video data is stored to a second memory location associated with the second source. This current set of digital video data is then identified as the representative video data for the second source.