Logo Pattern Matching for E-commerce Infringement Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
E-commerce platforms face challenges in accurately identifying and reducing infringing product images due to limitations in existing detection methods, such as rule-based systems and similarity matching, which lead to poor detection robustness and high recall rates, resulting in legal risks and increased manual auditing burdens.
Innovation Solution
A system and method that utilize a logo pattern library to calculate the matching degree of product images with logo patterns, selecting the logo pattern with the greatest matching degree and determining image infringement if the threshold is met, thereby improving detection accuracy and reducing false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If rule-based detection method is used to detect keywords in product data, then detection speed is improved, but detection accuracy deteriorates due to incomplete rules and seller evasion tactics
Solution Approach 1:
The patent segments the product image into multiple local regions and performs logo pattern matching on each region independently. This allows the system to maintain high detection speed through automated processing while improving accuracy by examining specific areas where logos are likely to appear, rather than relying on incomplete keyword rules.
Solution Approach 2:
The patent introduces logo pattern templates as an intermediary between the product image and the detection decision. These templates serve as reference standards that mediate the comparison process, enabling more accurate detection than direct keyword matching while maintaining automated processing speed.
2Reliability
If similarity matching method is used to compare product images with pre-collected images, then detection comprehensiveness is improved, but false positive rate increases due to variations in shooting angles and lighting
Solution Approach 1:
The patent divides the image comparison task into segmented logo pattern recognition within local regions. By focusing on specific logo patterns rather than overall image similarity, the system achieves comprehensive detection while reducing false positives caused by variations in shooting angles and lighting conditions that affect overall image appearance but not the logo itself.
Solution Approach 2:
The patent applies local quality by performing detection on specific local regions of the image where logos are expected to appear, rather than analyzing the entire image. This localized approach maintains detection comprehensiveness for logo-containing products while reducing false positives from global image variations in lighting and angle.
3Measurement precision
If exhaustive keyword rules are pre-configured for rule-based detection, then detection accuracy is improved, but system complexity increases and rules cannot be maintained
Solution Approach 1:
The patent uses pre-collected logo pattern templates as copies of authentic logos as reference standards. Instead of maintaining complex keyword rules, the system copies actual logo appearances and uses them for pattern matching, simplifying the system while improving detection accuracy through visual pattern recognition.
Solution Approach 2:
The patent changes the detection parameter from text-based keyword matching to image-based pattern matching. This parameter change reduces system complexity by eliminating the need for maintaining exhaustive keyword rules while improving detection accuracy through direct visual comparison with logo templates.
4Adaptability or versatility
If sellers use different shooting angles to avoid detection, then seller adaptability is improved, but detection robustness deteriorates with traditional similarity matching
Solution Approach 1:
The patent focuses detection on local logo regions rather than overall image characteristics. This local approach maintains detection robustness against seller adaptability tactics like changing shooting angles, because the logo pattern itself remains recognizable despite variations in image perspective and lighting conditions.
Data Source
AI summary
The present invention provides a method and device for detecting the presence of logo patterns in product images. The method comprises the steps of retrieving one or more logo patterns from a logo pattern library, each of the one or more logo patterns identifying a respective product brand; retrieving a product image; calculating for each of the one or more logo patterns, a matching degree, the matching degree representing a degree to which each logo pattern is present in the product image; selecting as a selected logo pattern, the logo pattern having the greatest calculated matching degree; and determining that the product image includes the selected logo pattern if the matching degree of the selected logo pattern is greater than a first threshold.