Video Display Apparatus Person Scene Detection Database
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video display apparatuses lack an efficient method for extracting and displaying thumbnail images from recorded video content, making it difficult for users to intuitively identify specific individuals within the content.
Innovation Solution
A video display apparatus with a person and face database that detects scenes featuring specific individuals by using face feature data, allowing users to select and display associated thumbnail images, and register person names with their corresponding face features for easy retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If thumbnail images are extracted from recorded video content using conventional methods, then the video content can be displayed in list form, but the user cannot intuitively identify specific individuals within the content
Solution Approach 1:
The patent segments the video content by detecting and extracting specific scenes containing target persons. The person video scene detecting module divides the continuous video stream into discrete scenes where specific individuals appear, allowing selective extraction and display of relevant thumbnails rather than displaying all content uniformly.
Solution Approach 2:
The patent introduces a person and face database as an intermediary component that stores face feature amounts and person names. This database mediates between the video content and the user by enabling face recognition and person identification, thereby preserving person identification information that would otherwise be lost in conventional thumbnail extraction.
2Measurement precision
If face feature data is used to detect scenes featuring specific persons, then person identification accuracy is improved, but device complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-storing face feature amounts and person names in the person and face database before video content analysis. This pre-processing step enables efficient person identification during video playback without requiring complex real-time analysis, thereby improving detection accuracy while managing system complexity.
Solution Approach 2:
The patent creates a simplified representation (copy) of person identification by storing face feature amounts as numerical data in the database. Instead of storing actual video frames or complex image data, the system uses compressed feature vectors that capture essential facial characteristics, reducing storage requirements and processing complexity while maintaining identification accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
According to one embodiment, a video display apparatus includes a database, a selecting module, a person video scene detecting module, and a video displaying module. The database includes data of a name of a person and a feature amount on an image of the person. The selecting module is configured to select a name of a specific person. The person video scene detecting module is configured to detect a scene including the person of the selected name from a video content by using the database. The video displaying module is configured to display at least one image of moving picture or stationary picture indicating the detected scene in the video content in a list form.