Feature extraction and automatic annotation of flash illuminated video data in unconstrained video streams

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention provides a method for automatically detecting and annotating flash illumination events in unconstrained video streams, such as those found in user-generated content. By breaking each frame into small segments and analyzing changes in their brightness over time, it can efficiently and accurately identify both whole-frame and localized flashes, even when only part of an image is affected. The method supports automatic classification and tagging of video scenes based on the presence of flash illumination, reducing errors found in previous techniques and enabling smarter content management.

Use CasesContent extracted from patent full text and abstract with AI.

• Automatic annotation of large video collections to indicate the presence of flash events, facilitating search and retrieval.
• Improving shot-boundary detection in video editing workflows, distinguishing actual scene cuts from flash events.
• Enhancing content moderation or curation on video-sharing platforms by flagging flash-illuminated scenes, which might signal highlights or events of interest.
• Supporting forensic analysis where detection of camera flashes could be relevant (e.g., in surveillance or incident reviews).
• Ranking or filtering social event videos for importance based on the frequency of flash events, useful for media sorting or highlights assembly.
• Automatically generating video metadata for indexing, accessibility, or archiving purposes.

BenefitsContent extracted from patent full text and abstract with AI.

• Accurately detects flash illumination even when affecting only small regions of a video frame, unlike previous methods focused on whole-frame changes.
• Reduces false positives by distinguishing flashes from similar luminance changes, such as cuts or fades.
• Enables fine-grained, automatic annotation and classification of video content, which supports efficient indexing and retrieval.
• Adaptable to various lighting conditions using histogram equalization and robust processing steps.
• Allows for efficient processing via dimension reduction (e.g., principal component analysis), making the method scalable for large databases.
• Applicable to unconstrained, real-world user-generated content, not just professionally produced or well-lit videos.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

A method of processing video data (1), comprises the steps of extracting the video data (1) into a plurality of frames (2), and analysing the plurality of frames (2) for detecting a flash illumination in the video data (1). Therein, analysing the plurality of frames (2) comprises the steps of partitioning each frame of the plurality of frames (2) into a plurality of segments (31) of a predefined size and calculating an average segment luminance component (41) for each segment of the plurality of segments (31) of each frame of the plurality of frames (2). By partitioning the video frames (2) into smaller segments (31) and subsequently using the transition in luminance component in each segment, an efficient and fine-tuned detection of flash illumination in the video data (1) is possible. Thereby, even flash illumination not affecting the whole frame can be detected which can be of decisive importance. Thus, when applying the method according to the invention flash detection in unconstrained user generated video clips or scenes can be significantly improved and accordingly the video clip and its frames can efficiently and sophistically be classified or rated as containing flash-illumination or not. Also, when detecting frames (2) that contain scenes with flash illumination according to the invention, the risk of incorrect recognition of flash illuminated scenes, e.g., by recognising cut points as flash illuminations, of false alarm can be reduced.