Real Face Recognition via PDAM Illumination and Fourier Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current face recognition systems are vulnerable to impostors using fake faces, and are affected by ambient lighting and resolution issues, leading to incorrect identification of real human faces as fake due to increased low frequency components.
Innovation Solution
The implementation of a Point Divid Arithmetic Mean (PDAM) Illumination Treatment followed by Fourier transformation to normalize illumination and distinguish real human faces from fake ones by analyzing high frequency components using a preset classification threshold and Support Vector Machine (SVM) classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional face recognition methods are used to analyze low frequency components, then the system can process face images, but the recognition accuracy deteriorates under varying ambient lighting conditions
Solution Approach 1:
The patent applies PDAM illumination treatment as a preliminary step before frequency analysis. This pre-processing normalizes the illumination of the face image by dividing each pixel value by the arithmetic mean of all pixel values, thereby eliminating the harmful effect of ambient lighting variations before the actual recognition process begins
Solution Approach 2:
The patent extracts and analyzes only the high frequency components of the face image after illumination normalization, while discarding the low frequency components that contain illumination information. This extraction of specific frequency components allows the system to focus on texture and structural features that are invariant to lighting conditions
2Adaptability or versatility
If face recognition systems accept images with varying resolutions, then the system is more versatile, but the detection accuracy deteriorates due to resolution-related distortions
Solution Approach 1:
The patent transforms the face image from spatial domain to frequency domain using Fourier transformation. This parameter change allows the system to analyze face features in terms of frequency components rather than spatial resolution, making the recognition process invariant to image resolution variations while maintaining high detection accuracy
3Measurement precision
If the system performs comprehensive frequency analysis on all image components, then the recognition thoroughness is improved, but the processing time increases
Solution Approach 1:
The patent extracts only the high frequency components of the face image for recognition analysis, deliberately excluding the low frequency components. This selective extraction reduces the amount of data that needs to be processed while maintaining recognition thoroughness, as the high frequency components contain the essential texture and structural information needed for accurate face recognition
Data Source
AI summary
A method for real face image recognition may include obtaining, by at least one processor, an human face image from an original image; obtaining, by at least one processor, a first image by executing a Point Divid Arithmetic Mean Illumination Treatment on the human face image; executing, by at least one processor, a Fourier transformation on the first image and obtaining, by at least one processor, the transformed value of each pixel of the first image; determining, by at least one processor, whether the human face image is a real human face according to the transformed value of each pixel of the first image and the preset classification threshold.