Patient Identification With Reference Templates for Waiting Rooms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face detection techniques struggle to accurately identify individual patients in digital images, particularly in varying conditions such as different poses, occlusions, and lighting, making it difficult to monitor patient conditions effectively in waiting rooms or other environments.
Innovation Solution
A system utilizing vital sign acquisition cameras and machine learning models, including geometric normalization and trained models, to match subject reference templates with detected faces in digital images, enabling contactless vital sign acquisition and patient monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If face detection techniques are used to identify all faces in digital images, then the system can detect multiple subjects, but it cannot distinguish which faces belong to patients and which belong to others
Solution Approach 1:
The system captures reference images of patients during intake before they enter the waiting room. These reference images are stored and later used for comparison to identify patients in monitoring images, enabling distinction between patients and other individuals in the scene
Solution Approach 2:
The system creates digital copies of patient faces in the form of reference images during intake. These copied images are then used for identification purposes by comparing them against images captured in the waiting room, allowing accurate patient identification without requiring real-time interaction
2Ease of operation
If cameras capture subjects in natural settings, then the system can monitor patients unobtrusively, but subjects are not likely to be looking at the camera making face detection difficult
Solution Approach 1:
Reference images are captured during intake when patients are instructed to face the camera directly. This preliminary capture of frontal views provides robust templates for later identification even when patients are not looking at the monitoring camera in the waiting room
Solution Approach 2:
The system transforms detected faces in monitoring images to match the orientation and characteristics of reference images. By adjusting parameters such as face orientation, scale, and lighting conditions, the system enables accurate comparison even when subjects are not looking at the camera
3Reliability
If the system monitors all patients continuously, then patient conditions can be tracked, but it requires accurate identification of each patient in the scene
Solution Approach 1:
The system uses copied reference images of each patient's face as a template for identification. By comparing the detected face in monitoring images against the stored reference copy, the system can reliably identify specific patients and track their conditions continuously
Solution Approach 2:
The system implements a feedback mechanism where detected faces are compared against reference images, and the comparison results feed back into the identification process. This iterative comparison and matching process improves identification precision by confirming matches through multiple verification steps
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed techniques relate to identifying subjects in digital images. In some embodiments, intake digital images (404) are acquired (1002) that capture a first subject. A subset of the intake digital images is selected (1004) that depict multiple different views of the first subject's face. Based on the selected subset of intake digital images, first subject reference templates are generated and stored in a subject database (412). Later, a second subject is selected (1008) for identification within an area. Associated second subject reference templates are retrieved (1010) from the subject reference database. Digital image(s) (420) that depict the area are acquired (1012). Portion(s) of the digital image(s) that depict faces of subject(s) in the area are detected (1014) as detected face image(s). A given detected face image is compared (1016) to the second subject reference templates to identify the second subject (1018) in the digital image(s) that capture the area.