Video Frame Selection Using Behavior Information Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video photographing methods face challenges in accurately selecting video frames, particularly when the photographed object is moving, as they rely on social network information scores, leading to low accuracy in frame selection.
Innovation Solution
A method and apparatus that acquire behavior information of a photographed person when a camera is turned on, determining if specific behavior information requires a photo, and configuring a target video frame based on this information, which can include expression, voice, or action behavior, and optionally performing resolution processing and selecting frames according to a photo selection policy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If social network information scores are used to select video frames, then photo generation is automated, but frame selection accuracy decreases
Solution Approach 1:
The patent segments the frame selection process into multiple independent criteria: behavior information detection (expressions, actions, voice), social network information scoring, and photo selection policy evaluation. Each criterion operates separately and contributes to the final selection, allowing automated processing while maintaining accuracy through multi-dimensional assessment.
Solution Approach 2:
The patent changes the selection parameters from relying solely on social network information scores to incorporating behavior information parameters such as expression detection, action recognition, and voice analysis. This parameter expansion enables more accurate frame selection while preserving automation through systematic evaluation of multiple parameters.
2Quantity of substance
If multiple video frames are captured for moving objects, then capture coverage is improved, but selection accuracy decreases
Solution Approach 1:
The patent implements feedback mechanisms where behavior information from captured frames is continuously analyzed to guide subsequent frame selection. The system uses detected expressions, actions, and voice patterns as feedback to identify and select the most relevant frames from the captured sequence, ensuring high accuracy even when multiple frames are available.
Solution Approach 2:
The patent applies dynamic selection criteria that adapt to the content of each video frame. Rather than static selection, the system dynamically evaluates behavior information in each frame and adjusts selection based on detected expressions, actions, and voice patterns, enabling accurate selection from multiple captured frames of moving objects.
3Measurement precision
If behavior information analysis is added to frame selection, then selection accuracy improves, but system complexity increases
Solution Approach 1:
The patent implements a universal photo selection apparatus that handles multiple functions: behavior information detection (expressions, actions, voice), social network information processing, and frame selection. This multi-functional system consolidates diverse processing tasks into a single integrated platform, managing complexity through unified architecture while maintaining high selection accuracy.
Solution Approach 2:
The patent performs preliminary analysis of behavior information during video capture and preprocessing stages. By detecting expressions, actions, and voice patterns in advance and pre-evaluating frame candidates, the system reduces the computational burden during final selection, managing complexity through staged processing while preserving accuracy.
Data Source
AI summary
Embodiments of the present disclosure disclose a video photographing processing method and apparatus, and relate to the field of information technologies. The embodiments of the present disclosure are applicable to automatic generation of a photo during video photographing. Various embodiments provide a video photographing processing method, including acquiring at least one piece of behavior information of a photographed person when a camera is turned on; determining whether behavior information for which a photo needs to be generated exists in the at least one piece of behavior information; and if the behavior information for which a photo needs to be generated exists in the at least one piece of behavior information, configuring a target video frame as a photo that needs to be generated.