Video Object Search via Dual-List Metadata Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video image processing techniques face inefficiencies in detecting specific objects due to increased metadata, leading to prolonged processing times when searching through large datasets.
Innovation Solution
An information processing apparatus with a first search unit for registering and identifying features in a registration list, a generation unit for creating separate lists for detected and undetected objects, and a second search unit for rapid target object identification, optimizing data storage and retrieval to enhance search speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-time search processing is performed on monitored video images, then detection capability is improved, but metadata storage increases and post-search processing time increases
Solution Approach 1:
The patent segments the search process into two distinct phases: real-time search during video monitoring and post-search processing on stored metadata. By dividing the originally unified search function into separate real-time and retrospective components, the system can maintain detection capability while reducing the time required for post-search operations through optimized metadata structures.
Solution Approach 2:
The patent performs preliminary extraction and storage of metadata during real-time video processing, preparing detection results in advance for future post-search operations. This preliminary action converts raw video data into structured metadata with detection timestamps and object information, significantly reducing the computational burden and time required during subsequent post-search processing.
2Reliability
If real-time search processing is performed on monitored video images, then detection capability is improved, but metadata storage increases
Solution Approach 1:
The patent extracts only the essential detection-related metadata (object type, detection timestamp, confidence level) from the video processing stream, separating this critical information from the full video data. This extraction approach maintains detection capability while storing only the necessary metadata elements, preventing uncontrolled storage growth.
Solution Approach 2:
The patent applies different storage strategies to different types of data: full-resolution video frames are stored selectively based on detection events, while compressed metadata is continuously stored. This local quality differentiation optimizes storage resource allocation by maintaining high fidelity where needed (detection moments) and using efficient compression where appropriate (metadata records).
Data Source
AI summary
An information processing apparatus includes a first search unit configured to search for a feature of an object extracted from a video image in a registration list in which a feature indicating a predetermined object to be detected and identification information for identifying the predetermined object are registered, a generation unit configured to generate a first list in which at least the ID information about the predetermined object corresponding to the extracted object is registered in a case where the feature of the extracted object is detected in the registration list and generate a second list in which the feature of the extracted object is registered in a case where the feature of the extracted object is not detected in the registration list, and a second search unit configured to search for a target object designated by a user in the first list or the second list.