Image Encoding Privacy Filter with Scattered Color Revision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding technologies face challenges in maintaining privacy while allowing for potential identification of objects in encoded video streams, as privacy filters distort images, making it difficult to extract original color information for later analysis.
Innovation Solution
A method that pre-processes images using a privacy filter to distort them, followed by color revision of scattered areas to represent original colors, allowing for later extraction of color information without compromising privacy, enabling identification of objects like persons or vehicles for investigation purposes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a privacy filter is applied to distort images, then privacy is achieved and identification of persons/vehicles is prevented, but extraction of original color information becomes difficult
Solution Approach 1:
The image is divided into multiple scattered areas, where only specific regions have their colors revised to represent original colors, while other regions remain distorted by the privacy filter. This segmentation allows selective preservation of color information in certain areas without compromising privacy protection in the entire image.
Solution Approach 2:
Different regions of the image are treated differently: some scattered areas maintain original color information for extraction purposes, while the rest of the image undergoes full privacy distortion. This local differentiation enables simultaneous achievement of privacy protection and color information preservation where needed.
2Loss of information
If all pixel colors are revised to represent original colors, then color information extraction is enabled, but privacy protection is compromised
Solution Approach 1:
Instead of revising all pixel colors, only colors in scattered areas are revised to represent original colors. This selective approach enables color information extraction while maintaining privacy protection in the majority of the image regions that remain distorted.
Solution Approach 2:
The patent applies color revision only partially to scattered areas rather than universally to all pixels. This partial action is sufficient for enabling color extraction when needed, while avoiding the excessive action of revising all colors which would compromise privacy.
3Adaptability or versatility
If scattered areas are introduced to preserve color information, then identification capability is enabled, but image processing complexity increases
Solution Approach 1:
The image processing is segmented into distinct operations: privacy filtering and selective color revision in scattered areas. This segmentation allows the complex task of maintaining both privacy and identification capability to be broken down into manageable processing stages.
Solution Approach 2:
The patent changes parameters selectively - applying color revision only to scattered areas rather than all pixels. This parameter change approach enables identification capability enhancement while controlling processing complexity by limiting the scope of color modification operations.
Data Source
AI summary
An image processing device a camera and a method for of encoding images captured by a camera are disclosed. For each image of an image sequence captured by the camera, the image is pre-processed by filtering the image by applying a privacy filter, the privacy filter being configured to distort the image in such a way that privacy is achieved in the filtered image, and, for at least a subset of the filtered images, by colour revising the filtered image by changing colour of pixels of a plurality of scattered areas of the filtered image such that a respective colour of one or more pixels of each area of the plurality of scattered areas represents one or more original colours of one or more pixels before filtering at a location of that area in the filtered image. The pre-processed images are then encoded into an encoded video stream.