Multi-Camera Face Authentication With Conversion Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices with multiple cameras having different characteristics face challenges in maintaining consistent face authentication reliability due to varying angle of view, resolution, and noise characteristics, leading to inconsistent performance across cameras.
Innovation Solution
An electronic device and method that utilize multiple cameras by generating conversion filters to normalize image characteristics, creating multiple templates, and adjusting priority based on matching scores to enhance face authentication reliability and user convenience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If face authentication is performed using a second camera with different characteristics, then user convenience is improved (multiple cameras can be used), but authentication reliability deteriorates due to inconsistent image characteristics
Solution Approach 1:
The patent applies parameter changes by adjusting image characteristics through conversion filters. The system transforms images from different cameras to have unified characteristics (resolution, angle of view, noise levels) by modifying parameters such as pixel density, filtering strength, and geometric transformation, enabling reliable authentication across multiple cameras
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of conversion filters and template adjustment layers. These intermediaries translate between the different characteristic spaces of multiple cameras, allowing images from any camera to be accurately compared against standardized templates
2Adaptability or versatility
If multiple cameras with different characteristics are used for face authentication, then device versatility is improved, but measurement precision deteriorates due to varying resolution and noise characteristics
Solution Approach 1:
The system changes image parameters through conversion filters to standardize resolution and noise characteristics. High-resolution camera images are downsampled with appropriate filtering, while low-resolution images are upscaled with noise reduction, ensuring all images meet the precision requirements for accurate feature extraction
Solution Approach 2:
The patent performs preliminary image processing and characteristic normalization before face feature extraction. By pre-adjusting images to unified characteristics and generating adaptive templates in advance, the system ensures that subsequent measurements and comparisons are performed on equally standardized data
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Various embodiments of the disclosure relate to an electronic device and a method for performing face authentication by using multiple cameras. The method of the electronic device may include an operation of receiving a request for registration of face information through a first camera device having a first characteristic, an operation of acquiring, based on the request, multiple first images through the first camera device, an operation of performing filtering for each of the multiple first images to generate multiple second images reflecting a second characteristic of the second camera device, an operation of generating multiple first templates based on the multiple first images, an operation of generating multiple second templates based on the multiple second images, and an operation of storing the multiple first templates and the multiple second templates in the memory, thereby completing the registration of the face information. The disclosure may include various other embodiments.