Phase Congruency Edge Detection for Varying Contrast
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing edge detection methods in image processing face challenges in accurately detecting edges in images with varying contrast, brightness, and intensity, particularly in scenarios like transit systems where distinguishing between similar objects or individuals is difficult.
Innovation Solution
The method employs phase congruency values and criteria to categorize pixels as edge or non-edge pixels, using a combination of phase congruency components and thresholds, and applies hysteresis to enhance edge detection, making it less sensitive to contrast and brightness variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional edge detection methods are used, then the process is simple and fast, but accuracy deteriorates in images with varying contrast and brightness
Solution Approach 1:
The patent changes the detection parameter from intensity-based to phase-based. By computing phase congruency values from Fourier components, the method detects edges based on phase relationships rather than intensity gradients, making it invariant to contrast and brightness variations while maintaining detection accuracy across diverse imaging conditions
Solution Approach 2:
The patent substitutes the traditional intensity-gradient-based mechanical detection system with a phase-congruency-based system. This replacement involves transforming image data into the frequency domain, computing phase angles of Fourier components, and detecting edges where phases are congruent, thereby eliminating sensitivity to intensity variations
2Reliability
If phase congruency method is applied, then accuracy in varying contrast conditions improves, but computational complexity increases
Solution Approach 1:
The patent performs preliminary Fourier transformation of the image data before edge detection. By pre-computing the Fourier components and their phases, the system prepares the data in advance, allowing the actual edge detection to proceed efficiently through simple phase comparison operations rather than complex real-time calculations
Solution Approach 2:
The patent segments the edge detection task into distinct computational steps: Fourier transformation, phase angle computation, phase congruency evaluation, and edge classification. This segmentation allows each step to be optimized independently and enables the system to handle complex images through modular processing
3Adaptability or versatility
If hysteresis is applied to edge detection, then sensitivity to contrast variations decreases, but additional processing steps are required
Solution Approach 1:
The patent implements hysteresis feedback by using two threshold values (first and second thresholds) for phase congruency evaluation. Pixels are classified as edges only when their phase congruency value exceeds the higher first threshold, while pixels with values between thresholds are classified as non-edges. This feedback mechanism filters out spurious edges caused by contrast variations while maintaining true edge detection
Data Source
AI summary
A method executed by a computer system for detecting edges comprises receiving an image comprising a plurality of pixels, determining a phase congruency value for a pixel, where the phase congruency value comprises a plurality of phase congruency components, and determining if the phase congruency value satisfies a phase congruency criteria. If the phase congruency value satisfies the phase congruency criteria, the computer system categorizes the pixel as an edge pixel. If the phase congruency value does not satisfy the phase congruency criteria, the computer system compares a first phase congruency component of the plurality of phase congruency components to a phase congruency component criteria. If the first phase congruency component satisfies the phase congruency component criteria, the computer system categorizes the pixel as an edge pixel, and if the first phase congruency component does not satisfy the phase congruency component criteria, categorizes the pixel as a non-edge pixel.