Face Image Prioritization via Quality Analysis Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video surveillance-based facial recognition systems face increased processing complexity due to high-resolution images and mechanical limitations of motor-driven cameras, leading to slower identification and location of individuals, as well as excessive data processing from multiple frames.
Innovation Solution
Implementing face image prioritization through quality analysis filters that assess factors like face size, brightness, sharpness, and overall image quality, allowing only high-quality images to be processed by the facial recognition program, and delaying processing until optimal image quality is achieved, thereby reducing processing load and conserving resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution images are used for facial recognition, then identification accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent segments the image processing task by first detecting face regions in low-resolution images, then selectively processing only those regions in high-resolution images. This divides the complex high-resolution processing into manageable segments based on detected face locations, reducing overall processing complexity while maintaining identification accuracy.
Solution Approach 2:
The patent applies different processing qualities to different parts of the image: low-resolution processing for the entire scene and high-resolution processing only for detected face regions. This local quality approach ensures high identification accuracy for faces while avoiding the complexity of processing entire high-resolution images.
2Measurement precision
If multiple high-resolution images are acquired to locate individuals, then location accuracy is improved, but data processing volume increases
Solution Approach 1:
The patent extracts face regions from full high-resolution images after initial detection in low-resolution images. By taking out only the relevant face portions for further processing, the system maintains location accuracy while significantly reducing the volume of data that requires intensive processing.
Solution Approach 2:
The patent performs preliminary face detection in low-resolution images before acquiring or processing high-resolution images. This preliminary action identifies which regions require high-resolution processing, preventing unnecessary processing of entire high-resolution images and reducing overall data processing volume.
3Manufacturing precision
If the higher resolution camera articulates to acquire images of located individuals, then image quality is improved, but response time increases due to mechanical limitations
Solution Approach 1:
The patent performs preliminary face detection and quality assessment using low-resolution images before the high-resolution camera articulates. This preliminary action allows the system to identify which individuals require high-resolution imaging and prepare processing pipelines in advance, reducing the effective response time despite mechanical articulation delays.
Solution Approach 2:
The patent maintains continuous processing by preparing quality analysis filters and processing pipelines in advance while the high-resolution camera articulates. This continuity ensures that when high-resolution images arrive, processing can begin immediately without waiting for camera movement to complete, effectively reducing response time.
4Reliability
If a large number of images are processed for facial recognition, then detection completeness is improved, but processing complexity increases
Solution Approach 1:
The patent segments the large set of images into groups based on detected face regions and quality metrics. By processing images in segmented groups rather than all at once, the system maintains detection completeness while managing processing complexity through divided, manageable tasks.
Solution Approach 2:
The patent applies partial processing to images that fail quality thresholds, using simplified analysis rather than full facial recognition processing. This partial action approach maintains detection completeness by still examining all images while reducing overall processing complexity through selective application of processing intensity.
Data Source
AI summary
Methods, machine-readable media, and devices for face image prioritization based on face quality analysis are described herein. For example, one or more embodiments include detecting a facial image in an image that has been acquired by a camera that monitors a scene, passing the facial image through a number of quality analysis filters that include a number of quality analysis factors, wherein processing complexity associated with the number of quality analysis factors increases consecutively, and submitting the facial image to a facial recognition program upon a determination that the facial image has passed the number of quality analysis filters.