Contour-Based Privacy Masking for Image De-identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional privacy masking technologies inadvertently lose pixel information from regions that do not require de-identification, deteriorating the image value and failing to efficiently share contour-based privacy region settings between masking and unmasking apparatuses.
Innovation Solution
A contour-based privacy region masking apparatus and method that detects object contours in images, separates privacy and background regions, generates a mask map with pixel values set to '1' for privacy regions and '0' for backgrounds, and uses Run-Length Encoding for data lossless compression to transmit masking region descriptors, enabling accurate de-identification and reconstruction of original images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If rectangular or circular masking regions are used for privacy masking, then the implementation is simple, but pixel information in non-privacy regions is unnecessarily lost
Solution Approach 1:
The patent segments the image into precise privacy regions and non-privacy regions using contour detection. Instead of applying a single rectangular or circular mask to the entire privacy area, the system divides the masking operation into precise contour-based regions, allowing selective masking only where needed while preserving information in non-privacy areas.
Solution Approach 2:
The patent applies different quality treatments to different regions of the image. Privacy regions detected through contour analysis receive masking treatment, while non-privacy regions maintain their original pixel information quality. This local differentiation ensures that only necessary areas are masked, preventing unnecessary information loss.
2Loss of information
If contour-based privacy region detection is used, then pixel information loss is minimized, but the device complexity increases
Solution Approach 1:
The patent introduces a mask map as an intermediary data structure that stores contour information in a simplified format. The contour detection algorithm generates this mask map that represents privacy regions, which then serves as a intermediary for the masking operation. This separates the complex detection phase from the simpler masking phase, managing overall system complexity.
Solution Approach 2:
The patent changes the parameter representation from complex contour geometry to a simplified mask map format with pixel values of '1' for privacy regions and '0' for non-privacy regions. This parameter transformation simplifies subsequent processing while preserving the essential contour-based region information needed for precise masking.
3Reliability
If mask map information is transmitted between masking and unmasking apparatuses, then accurate reconstruction is enabled, but data transmission volume increases
Solution Approach 1:
The patent transmits a simplified copy of the mask map information rather than the entire original image or complex contour data. The mask map contains only essential binary information ('1' for privacy regions, '0' for non-privacy regions), creating a compact representation that enables accurate reconstruction without transmitting excessive data.
Solution Approach 2:
Instead of transmitting the original image data or complex contour descriptions, the patent inverts the approach by transmitting a difference map (mask map) that indicates only where changes occurred. This inverted data representation method reduces transmission volume while preserving reconstruction accuracy.
Data Source
AI summary
Disclosed herein are a contour-based privacy region masking apparatus and method. The contour-based privacy region masking apparatus includes a memory for storing at least one program, and a processor for executing the program, wherein the program may perform detecting a contour of an object that is a privacy protection target in an original image, setting a masking region based on the detected contour, and de-identifying the masking region in the original image.