Video Processing System Identifying Areas of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Capturing and processing video of fast-paced and unpredictable events, such as sporting events, is challenging due to the difficulty in identifying and focusing on areas of interest within the vast field of view captured by multiple cameras.
Innovation Solution
A system and method that determine movement profiles of movable objects within the video data, correlate these profiles with a venue-specific template to assign interest ratings, and identify a sub-portion of the area based on these ratings, allowing for the generation and output of a cropped video with reduced resolution and bitrate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple cameras with wide field of view are used to capture the entire area, then the coverage of the area is improved, but the difficulty in identifying areas of interest increases
Solution Approach 1:
The patent segments the video data into multiple frames and identifies sub-portions of interest within each frame by analyzing movement profiles of objects. This allows the system to maintain wide area coverage while breaking down the complex task of identifying areas of interest into manageable segments, thereby resolving the contradiction between comprehensive coverage and detection difficulty.
Solution Approach 2:
The patent introduces movement profiles as an intermediary representation that bridges the gap between the complete video data and the identification of areas of interest. By creating a simplified model of object movement, the system can efficiently identify relevant areas without processing the entire video stream in detail, thus maintaining coverage while reducing detection complexity.
2Measurement precision
If the entire video data is processed to maintain high resolution, then the quality of video information is improved, but the bandwidth requirements increase
Solution Approach 1:
The patent extracts only the necessary sub-portions of video data that contain areas of interest, rather than processing and transmitting the entire video stream. By taking out only the relevant portions based on movement profile analysis, the system maintains high video quality for important areas while significantly reducing overall bandwidth requirements.
Solution Approach 2:
The patent applies different quality levels to different parts of the video data. High-resolution processing is applied locally to sub-portions identified as areas of interest, while lower quality or no processing is applied to areas without interest. This local quality approach maintains measurement precision where needed while reducing overall energy and bandwidth consumption.
Data Source
AI summary
A system and method for processing video is disclosed. The method includes receiving video of an area; determining a movement profile for each of a plurality of movable objects in the video; determining an interest-rating for each movable object of the plurality of movable objects based on a correlation between the movement profile of that object and a template; and identifying a sub-portion of the video based upon the interest-ratings.