Privacy-Aware Visual SLAM Map Updating With Modified Key Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visual-based SLAM technologies in spatial localization and tracking are hindered by the presence of private information, which increases computational complexity and burden, and necessitate encryption to prevent data breaches.
Innovation Solution
Implement a privacy content recognition model to identify and process privacy content in image frames, generating modified key frames with smoothed or blurred areas to exclude private information, thereby updating the environment map without using the private data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If visual-based SLAM technology is used for spatial localization and tracking, then localization and tracking performance is improved, but computational complexity and computational burden increase due to private information
Solution Approach 1:
The patent extracts and removes private information (personnel, documents, etc.) from image frames before processing. By identifying and extracting these privacy contents, the system reduces computational burden while maintaining localization accuracy, as the extracted private data no longer interferes with SLAM computations.
Solution Approach 2:
The patent segments the image frame into different regions, separating private information areas from public/environmental areas. This segmentation allows the system to process only the relevant environmental information for SLAM while excluding private content, thereby reducing computational complexity without sacrificing localization performance.
2Reliability
If private information is processed in SLAM system, then spatial localization and tracking can be performed, but data breaches may occur
Solution Approach 1:
The system extracts and removes private information from the processing pipeline before SLAM computation. By taking out personnel, documents, and other private data from image frames, the system eliminates data breach risks while preserving the ability to perform accurate spatial localization and tracking using only environmental features.
Solution Approach 2:
The patent converts the presence of private information from a harmful factor into a detection opportunity. By using privacy content recognition models to identify private data, the system transforms what would be a security risk into a feature that triggers selective processing, ultimately protecting against data breaches while maintaining SLAM functionality.
3Object-affected harmful factors
If encryption is applied to prevent data breaches, then data security is improved, but computational performance decreases
Solution Approach 1:
Instead of encrypting data to protect against breaches, the patent extracts and removes private information before processing. This approach eliminates the need for encryption operations, thereby maintaining high computational performance while still preventing data breaches by ensuring private data never enters the processing pipeline.
Solution Approach 2:
The system converts the security concern into a preprocessing step that identifies and removes private data before computation. This transforms the security problem from requiring encryption (which slows computation) to requiring selective data extraction (which maintains performance), achieving both security and speed.
Data Source
AI summary
The present disclosure provides data processing method and electronic device. The data processing method is applicable to the electronic device in an environment. The data processing method includes: obtaining a key frame; in a condition that the key frame comprises at least one privacy content, processing the key frame to generate a modified key frame, wherein the modified key frame comprises at least one image processing area corresponding to the at least one privacy content; and using the modified key frame to update a map of the environment.