Laboratory Image Masking for Confidential Object Privacy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Laboratory instruments capture images and videos that may expose confidential information, such as personal data or equipment details, compromising data privacy.
Innovation Solution
A method and system that applies a virtual masking object on or around predefined target objects in laboratory environments using image capturing devices, employing machine learning to identify and secure sensitive data by masking face regions or equipment with virtual curtains.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If images and videos are captured by the image capturing device, then operations of laboratory instruments are recorded, but confidential information such as personal information and equipment details are exposed
Solution Approach 1:
The system performs preliminary identification of confidential objects (face regions, ID cards, equipment) in the captured images before the images are fully processed or stored. By pre-identifying these sensitive regions using machine learning models, the system can apply masking treatments in advance, preventing confidential information from being exposed in subsequent processing, storage, or transmission stages.
Solution Approach 2:
The patent introduces a virtual masking object as an intermediary element between the confidential information and the external environment. This masking object (such as a virtual curtain or overlay) is superimposed on the identified confidential regions, acting as a mediator that blocks the exposure of sensitive data while allowing the rest of the image to remain visible and useful for operational purposes.
2Reliability
If virtual masking objects are applied on identified objects, then data privacy is protected, but image quality and completeness are reduced
Solution Approach 1:
The system applies the virtual masking object only to specific local regions of the image where confidential information is detected (such as face regions, ID cards, or equipment areas), rather than masking the entire image. This localized approach ensures that data privacy is protected in critical areas while preserving the quality and completeness of the rest of the image, allowing operational information to remain visible and useful.
3Measurement precision
If machine learning is used to identify target objects, then accurate detection of confidential data is achieved, but system complexity increases
Solution Approach 1:
The patent extracts and isolates the function of confidential object identification into a separate machine learning module that can be independently trained and optimized. By taking out this complex detection function as a standalone component (such as a pre-trained neural network model), the main laboratory instrument system doesn't need to handle the complexity of developing and managing the entire machine learning pipeline, reducing overall system complexity while maintaining high detection accuracy.
Data Source
Figure 1A~1B
Figure 2A
Figure 2B
AI summary
Disclosed subject matter relates to method and system for securing data of objects in a laboratory environment. An image capturing device configured in a laboratory instrument may capture images of plurality of objects in the laboratory environment. A processor in the laboratory instrument may identify one or more objects from the images matching with predefined target objects. The processor may apply virtual masking object on or around the identified objects to prevent exposure of data associated with the identified objects and thus provides data privacy.