| 摘要 |
<p>A method and system for augmenting a video with captions is presented. Caption sets for a video are generated along with timing tags, the captions associated with specific elements within the video. Preferably the video advertises a service or product, such as a vehicle, and the video may be standardised so a tag and caption will be timed with the same aspect of the product, e.g. a vehicle engine or interior etc. The caption data may be generated by querying a database(s) using search terms extracted from the video. For vehicle videos, a registration plate may be read by an automatic image analysis system (ANPR) and this ID used to gather make, model and specification data. The captions may be arranged in categorized sets. The tags may contain display duration, positioning data etc. for the captions. The captions may be overlaid on to the video images, or the captions may be displayed alongside the video. The video may be caused to pause when a caption is presented to allow the caption to be read.</p> |
