Obfuscation Network for Privacy-Preserving Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current security systems using cameras face limitations in tracking individuals while protecting private information, requiring complex obfuscation processes to anonymize images, which restricts effective incident analysis and data usability.
Innovation Solution
A method and device for tracking objects by obfuscating images using an obfuscation network that generates unidentifiable images to humans but identifiable by a learning network, allowing detection and tracking of obfuscated targets while matching consented non-obfuscated identification information with obfuscated tracking information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If original images are used for tracking, then tracking accuracy is improved, but identification information privacy is compromised
Solution Approach 1:
The patent introduces an obfuscation network as an intermediary between the original image and the tracking system. This network transforms original images into obfuscated images that preserve spatial-temporal features necessary for tracking while removing identifiable human features. The obfuscated images serve as a mediator that enables tracking functionality while protecting privacy, resolving the contradiction between tracking accuracy and privacy protection.
Solution Approach 2:
The patent applies parameter changes by transforming the visual parameters of images through the obfuscation network. The network modifies image parameters such as pixel values, color distributions, and feature representations to eliminate identifiable information while maintaining structural patterns needed for tracking. This allows the system to operate on transformed parameters that preserve utility while removing harmful identification capabilities.
2Object-affected harmful factors
If obfuscation is applied to protect privacy, then privacy protection is improved, but tracking capability deteriorates
Solution Approach 1:
The obfuscation network performs selective parameter changes that preserve critical tracking parameters while modifying identification parameters. It maintains spatial coordinates, movement patterns, and temporal sequences necessary for tracking, while altering facial features, clothing details, and other identifiable characteristics. This selective parameter transformation resolves the contradiction by preserving tracking-relevant parameters while changing privacy-relevant parameters.
Solution Approach 2:
The patent segments the image processing function into two distinct components: the obfuscation network that handles privacy protection, and the learning network that handles tracking detection. This segmentation allows each component to specialize - the obfuscation network focuses on privacy preservation while the learning network focuses on tracking accuracy, resolving the contradiction through functional separation.
3Object-affected harmful factors
If complex obfuscation operations are applied, then privacy protection is improved, but system complexity increases
Solution Approach 1:
The patent replaces complex manual or rule-based obfuscation operations with a learned neural network model. Instead of applying multiple separate obfuscation techniques (blurring, pixelation, masking), the system uses a single trained obfuscation network that automatically performs the necessary transformations. This substitution reduces operational complexity while maintaining or improving privacy protection effectiveness.
Solution Approach 2:
The obfuscation network is trained in advance on large datasets to learn the optimal transformation patterns for privacy protection. This preliminary training action allows the network to perform complex obfuscation tasks efficiently during actual operation without requiring real-time complex computations. The heavy lifting is done beforehand, reducing runtime system complexity.
4Loss of information
If original images are used for analysis, then data usability is improved, but identification information exposure increases
Solution Approach 1:
The obfuscated image serves as an intermediary data representation that preserves analytical value while eliminating identification information. The learning network can extract behavioral patterns, movement trajectories, and incident characteristics from obfuscated images with high usability, while the obfuscation ensures no identification information is exposed. This intermediary representation resolves the contradiction between data usability and information exposure.
Data Source
AI summary
A method for tracking one or more objects in a specific space is provided. The method includes steps of: (a) inputting original images of the specific space taken from camera to an obfuscation network and instructing the obfuscation network to obfuscate the original images to generate obfuscated images such that the obfuscated images are not identifiable as the original images by a human but the obfuscated images are identifiable as the original images by a learning network; (b) inputting the obfuscated images into the learning network, and instructing the learning network to detect obfuscated target objects, corresponding to target objects to be tracked, in the obfuscated images, to thereby output information on the obfuscated target objects; and (c) tracking the obfuscated target objects in the specific space by referring to the information on the obfuscated target objects.