Face Capture System for Moving Subjects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face recognition systems require individuals to stop and hold still for image capture, leading to slow and tedious processes, and often fail to capture high-quality images of moving persons.
Innovation Solution
A face recognition system that tracks and captures high-quality images of moving individuals using a guidance feature, such as a handrail or augmented reality projector, while employing image processing techniques to enhance image quality and a face detection engine to select the maximum quality image for matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system requires the candidate to stop and hold still for image capture, then the image quality is improved, but the processing time and user wait time increase significantly
Solution Approach 1:
The system performs preliminary actions by pre-positioning multiple cameras at different angles and preparing image processing algorithms before the candidate arrives. This allows the system to rapidly capture and evaluate multiple images simultaneously as the candidate walks through, eliminating the need for the candidate to stop while still ensuring high-quality image capture through pre-prepared multi-angle coverage
Solution Approach 2:
The system captures excessive images by using multiple cameras to take numerous photos from different angles and distances as the candidate walks through. This excessive capture ensures that at least one high-quality image is obtained without requiring the candidate to hold still, as the system processes multiple candidate images to select the best match
2Device complexity
If the system uses a single camera to capture the face image, then the device complexity is reduced, but the reliability of capturing high-quality images of moving persons deteriorates
Solution Approach 1:
The system segments the image capture function by using multiple cameras positioned at different locations and angles. Each camera captures images from its specific viewpoint, and the system processes these segmented image streams to select the highest quality image. This segmentation allows reliable capture of moving subjects without requiring a single complex camera system
Solution Approach 2:
The multiple cameras serve universal functions by collectively providing coverage that no single camera could achieve alone. They simultaneously perform detection, tracking, and image capture from multiple perspectives, making the system reliable for capturing moving persons while maintaining manageable device complexity through standardized camera units
3Measurement precision
If the system processes multiple images to select the maximum quality image, then the image selection accuracy is improved, but the computational time and processing complexity increase
Solution Approach 1:
The system performs preliminary quality assessment by evaluating basic image parameters such as face detection confidence, image sharpness, and proper framing before full processing. This preliminary filtering reduces the number of images requiring detailed analysis, maintaining high selection accuracy while reducing overall computational complexity and processing time
Data Source
Figure 1A~1D
Figure 2
Figure 3
AI summary
A fast face capture system and process for identifying an individual as the individual walks through a designated area is described. A set of raw images is streamed from at least one camera for detecting individuals entering the designated area. If a face is detected, face tracking is commenced to obtain a sequence of candidate images for each individual as the individual walks through the designated area. A maximum quality image is selected from the sequence of candidate images for each individual based on at least one quality metric, the elapsed time, and the quality select count in the sequence of the candidate images. The maximum quality image is submitted for matching with a verified image of the person. The invention has particular application to security check points for quickly matching the face of the moving person with a previously acquired and verified identity.