Embedded Face Recognition Hardware Using Parallel Histogram Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face recognition systems face challenges in real-time operation due to high system costs and large data throughput, particularly when using sequential methods on Personal Computers, which limits their effectiveness and increases system size.
Innovation Solution
A real-time face recognition apparatus utilizing a non-statistical method that includes a face detection unit, eye detection unit, facial feature extraction unit, database unit, and histogram matching unit, capable of parallel processing, employing Gabor filters and Multi-resolution Local Binary Patterns to generate and compare feature histograms, and outputting identification information efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional face recognition algorithms are implemented using software on Personal Computers, then face recognition accuracy can be achieved, but system cost increases and system size increases
Solution Approach 1:
The patent replaces the conventional software-based face recognition system running on Personal Computers with a hardware-based embedded system. This substitution transitions from a software-intensive approach to a dedicated hardware implementation, reducing system size and cost while maintaining recognition accuracy through specialized processing units designed for face recognition tasks.
Solution Approach 2:
The patent changes the operational parameters by implementing parallel processing capabilities in the embedded system, allowing simultaneous execution of multiple face recognition operations. This parameter change enables real-time processing and reduces the computational burden on individual processing units, thereby reducing system complexity while maintaining accuracy.
2Adaptability or versatility
If conventional face recognition algorithms are implemented using software on Personal Computers, then face recognition functionality can be provided, but system cost becomes high
Solution Approach 1:
The patent replaces expensive software-based implementations with a cost-effective embedded hardware system. This substitution eliminates the need for high-performance Personal Computers and complex software licenses, reducing system cost while maintaining full face recognition functionality through dedicated hardware processing units.
Solution Approach 2:
The patent uses standardized embedded system components and modular design patterns that can be replicated and manufactured at scale. By copying proven hardware architectures and processing algorithms into the embedded system, the patent reduces development costs and enables mass production, thereby reducing overall system cost while maintaining functionality.
3Device complexity
If sequential processing methods are used for face recognition, then system implementation is simpler, but real-time operation capability is lost due to large data throughput requirements
Solution Approach 1:
The patent implements a dynamic processing architecture that can adaptively allocate processing resources based on input data characteristics and throughput requirements. This dynamic approach allows the system to maintain simplicity in basic operations while enabling real-time processing through resource scaling and parallel execution when needed, resolving the contradiction between simplicity and real-time capability.
Solution Approach 2:
The patent segments the face recognition processing into multiple independent parallel processing units, each handling specific aspects of face analysis. This segmentation allows simultaneous processing of multiple faces or multiple processing stages, enabling real-time operation while keeping individual processing units simple and manageable.
4Power
If PC-based systems are used for face recognition, then computational power can be sufficient, but system size increases
Solution Approach 1:
The patent replaces general-purpose PC hardware with specialized embedded processing units optimized for face recognition tasks. This substitution achieves sufficient computational power through dedicated hardware accelerators and optimized processing pipelines while dramatically reducing the physical system size by eliminating unnecessary PC components and using compact embedded form factors.
Data Source
AI summary
Disclosed herein is a real-time face recognition apparatus and method. A real-time face recognition apparatus includes a face detection unit for detecting a face image by obtaining image coordinates of a face from an input image. An eye detection unit obtains image coordinates of both eyes in the face image. A facial feature extraction unit generates feature histogram data based on parallel processing from the face image. A DB unit stores predetermined comparative feature histograms. A histogram matching unit compares the histogram data generated by the facial feature extraction unit with the comparative feature histograms, and then outputting similarities of face images. The face recognition apparatus may be implemented as internal hardware in which a VGA camera and an exclusive chip interface with each other, thus remarkably reducing a system size and installation cost, and performing face recognition in real time without requiring additional equipment.