Dynamic Video Masking for Privacy and Detail Observation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance video processing systems fail to efficiently set masking and unmasking conditions based on the specific situation in videos, leading to inadequate privacy protection for various objects.
Innovation Solution
A video processing device and system that includes an object detector, event detector, masking target setter, and masking process controller, allowing users to set masking conditions for objects and unmasking conditions for specific events, enabling dynamic masking operations based on user inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If uniform masking operations are performed on all objects of a selected type, then privacy protection is ensured, but the ability to observe detailed status of specific objects is lost
Solution Approach 1:
The patent applies local quality by differentiating masking treatment between different spatial locations. Objects in the background area receive uniform masking for privacy protection, while objects in the foreground area (detected by event detector) remain unmasked to allow detailed observation. This resolves the contradiction by applying different masking qualities to different regions of the video frame.
Solution Approach 2:
The patent segments the video frame into multiple areas: foreground area containing objects of interest and background area containing other objects. By segmenting the processing approach, the system can apply uniform masking to background objects while preserving foreground objects, thus maintaining both privacy protection and detailed observation capabilities.
2Productivity
If masking conditions are set according to video situation, then efficient privacy protection is achieved, but system complexity increases
Solution Approach 1:
The system implements self-service by automatically detecting objects and events in the video, then autonomously determining which objects should be masked and which should remain unmasked. The event detector automatically identifies objects requiring detailed observation, and the masking controller automatically applies appropriate masking without requiring manual configuration, thus achieving high efficiency without increasing operational complexity.
Solution Approach 2:
The patent applies preliminary action by pre-configuring the system with masking rules and event detection criteria. The object detector and event detector are pre-programmed to automatically identify relevant objects and events, so when video processing occurs, the masking conditions are already established, enabling efficient real-time processing without complex runtime decision-making.
3Reliability
If all objects of selected type are set as masking targets, then privacy protection is comprehensive, but objects requiring detailed observation cannot be viewed
Solution Approach 1:
The patent applies dynamics by making the masking status of objects dynamic rather than static. Objects can transition between masked and unmasked states based on real-time event detection. When an object becomes the focus of an event, it dynamically transitions from masked to unmasked state, allowing detailed observation while maintaining comprehensive privacy protection for non-event objects.
Solution Approach 2:
The system implements feedback through the event detector that continuously monitors objects and provides feedback to the masking controller. When specific events are detected involving certain objects, this feedback triggers the controller to adjust masking status accordingly, ensuring that objects requiring detailed observation are unmasked while maintaining comprehensive privacy protection for other objects.
Data Source
AI summary
A video processing device includes: an object detector configured to detect an object included in a video so that a type of the object can be identified; an event detector configured to detect a focusing event which occurs to the detected object; a masking target setter configured to simultaneously set all objects of a selected type as masking targets, and to exclude from the masking targets, one or more objects to which occurrence of the selected focusing event has been detected by the event detector; and a masking process controller configured to generate a masking-processed video in which each video area which corresponds to the object set as the masking target by the masking target setter, has been changed to a masking image.