Video Surveillance Event Detection With Operator Feedback Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing behavior recognition systems in video surveillance lack the ability for real-time discriminator learning, leading to a decrease in analytical accuracy during operation due to the absence of operator intervention and support.
Innovation Solution
A video processing apparatus and system that includes a video analyzer, display controller, and learning data accumulator, allowing operators to set category information through a category setting screen, and accumulate this data for learning processing, thereby enhancing the system's analytical accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning is used for behavior recognition without operator intervention, then the system can operate automatically, but the analytical accuracy cannot be improved during actual operation
Solution Approach 1:
The system implements feedback by allowing operators to review detection results and provide correction inputs through the category setting screen. The learning data accumulator stores both original video data and operator-corrected category information, creating a feedback loop where operator insights continuously improve the detection model's accuracy while maintaining automatic operation.
Solution Approach 2:
The system enables self-service learning by automatically accumulating operator corrections and using them to retrain the detection model. The video analyzer performs learning processing using accumulated learning data without requiring manual reconfiguration, allowing the system to self-improve its analytical accuracy during actual operation.
2Measurement precision
If operator intervention is added for category setting, then the analytical accuracy can be improved, but the system complexity increases
Solution Approach 1:
The category setting screen serves multiple functions: it displays detection results for operator review, collects operator corrections as learning data, and manages category classifications. This multi-functionality reduces the need for separate interfaces or systems, minimizing added complexity while enabling accuracy improvement through operator intervention.
Solution Approach 2:
The learning data accumulator acts as an intermediary component that bridges automatic detection and operator input. It systematically stores and manages both original detection data and operator corrections, facilitating the transfer of knowledge from operators to the detection model without requiring direct complex interactions between operators and the learning algorithm.
3Adaptability or versatility
If learning data is accumulated from operator inputs, then the detection capability improves over time, but the data management complexity increases
Solution Approach 1:
The system merges the accumulation of video data and category information into a single learning data accumulator. This unified structure stores both types of data together with their associations, simplifying data management compared to separate systems while enabling comprehensive learning that improves detection capability across multiple event categories.
Solution Approach 2:
The system performs preliminary organization of learning data by systematically storing video data and category information together in the learning data accumulator before training processes occur. This preliminary data preparation and organization simplifies subsequent learning operations and enables efficient detection capability improvement without complex real-time data management during training.
Data Source
AI summary
A video processing apparatus includes a video analyzer that analyzes video data captured by a surveillance camera, detects an event belonging to a specific category, and outputs a detection result, a display controller that displays, together with a video of the video data, a category setting screen for setting a category of an event included in the video, and a learning data accumulator that accumulates, as learning data together with the video data, category information set in accordance with an operation by an operator to the category setting screen. The video analyzer performs learning processing by using the learning data accumulated in the learning data accumulator.