Per-pixel filtering for video privacy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in effectively filtering out specific objects or information from video transmissions to prevent the leakage of confidential information, which can compromise user privacy and security.
Innovation Solution
A per-pixel filtering method that involves generating pixel characterization vectors for each pixel in an image data frame, modifying pixel data corresponding to specific object labels, and synthesizing a modified image data frame that obfuscates or replaces the identified objects while leaving other pixels unmodified.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If video transmission is blocked entirely to prevent confidential information leakage, then information security is improved, but user experience and communication quality deteriorate
Solution Approach 1:
The patent segments the video feed into individual pixel-level processing units. Each pixel is analyzed independently to determine whether it contains confidential information, allowing selective filtering at the pixel level rather than blocking the entire video transmission. This enables partial transmission of video data while maintaining security.
Solution Approach 2:
The patent applies different processing qualities to different regions of the video feed. Pixels containing confidential information are filtered or obscured, while pixels containing non-sensitive information are transmitted normally. This creates a spatially varying quality output that maintains overall video usability while protecting specific sensitive areas.
2Reliability
If per-pixel filtering is implemented to selectively obfuscate sensitive objects, then information security is improved, but device complexity increases
Solution Approach 1:
The patent extracts and removes specific pixels or pixel groups from the video transmission stream that contain confidential information. By identifying and extracting only the sensitive portions at the pixel level, the system achieves targeted security filtering without requiring complex transformation of the entire video data stream.
Solution Approach 2:
The patent creates a filtered copy of the video feed where sensitive pixels are obscured or removed. This copied version is then transmitted instead of the original, allowing the transmitting device to maintain the full-quality local copy while sending a security-filtered version to remote devices.
Data Source
AI summary
Various implementations disclosed herein include devices, systems, and methods for per-pixel filtering. In some implementations, a method includes obtaining an image data frame. In some implementations, the image data frame includes a plurality of pixels. In some implementations, the method includes generating a respective pixel characterization vector for each of the plurality of pixels. In some implementations, each pixel characterization vector includes an object label indicating an object type that the corresponding pixel of the plurality of pixels represents. In some implementations, the method includes modifying corresponding pixel data of the plurality of pixels having a first object label. In some implementations, the method includes synthesizing a first modified image data frame that includes modified pixel data for the plurality of pixels having the first object label and unmodified pixel data for the plurality of pixels not having the first object label.