Product Authentication via Significant Point Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for authenticating luxury or technical products, such as watches, face challenges in distinguishing genuine from counterfeit items without altering their appearance or functionality, and struggle to accurately differentiate between relevant and irrelevant differences in images due to noise and natural changes over time.
Innovation Solution
The method involves extracting and comparing significant points from images of the product to be authenticated with those of a genuine product, using algorithms that are robust to noise and distortion, and aligning these points to determine authenticity, even when only a portion of the image is available, by calculating descriptors and applying transformations to align the images accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If image comparison is performed on raw image data, then comprehensive information is available for authentication, but processing demand and computational complexity increase significantly
Solution Approach 1:
The patent extracts significant points (key features) from the entire image data, isolating only the most relevant information for authentication. This extraction process removes unnecessary data while retaining the essential characteristics needed for accurate comparison, thereby reducing processing demand without compromising authentication reliability
Solution Approach 2:
The patent creates simplified representations (copies) of the original images in the form of significant points and descriptors. These point-based representations serve as compact substitutes for full image data, enabling efficient comparison while preserving the critical authentication information
2Reliability
If all differences between two images are considered, then potential authentication clues are not missed, but irrelevant differences due to noise and natural changes increase false positives
Solution Approach 1:
The patent applies different processing qualities to different parts of the image by identifying and weighting significant points differently. Critical features receive higher weight and more rigorous matching criteria, while less important areas are processed with lower scrutiny. This local differentiation allows the system to maintain high sensitivity to relevant differences while filtering out noise and natural variations in less critical regions
Solution Approach 2:
The patent changes the parameter of image representation from continuous pixel data to discrete significant points with associated descriptors and weights. This parameter transformation enables selective emphasis on authentication-critical features while naturally downplaying irrelevant variations, improving the system's ability to distinguish meaningful differences from noise
3Adaptability or versatility
If images captured under varying lighting conditions are compared, then real-world authentication flexibility is improved, but image degradation and comparison accuracy deteriorate
Solution Approach 1:
The patent performs preliminary processing to extract significant points and compute descriptors that are inherently more robust to lighting variations than raw pixel values. By preparing these lighting-invariant features in advance, the system maintains comparison accuracy even when images are captured under different lighting conditions, thus achieving both flexibility and precision
Data Source
Figure 1
Figure 2~10
Figure 3A
AI summary
A method of authenticating a product by taking an image of the product and comparing the image with a reference image of a genuine product taken previously to determine if the products in the two images are the same. The two images are captured under substantially similar conditions so that the two images are as similar as possible prior to the comparison. The two images are processed in order to calculate for each of them a list of significant points. The significant points are compared to determine a degree of correspondence between the significant points. An answer is output indicating the authenticity of the product based on the degree of correspondence. Some of the matching significant points may be used to define a common coordinate system for the two images. The two lists of significant points may be compared in this common coordinate system.