Leather product comparison and identification method and system based on image recognition

By collecting dynamic surface deformation videos of leather products, extracting multi-dimensional physical and mechanical features, and generating an adaptive identification model, the problems of interference from factors such as lighting and rigid decision-making in existing technologies are solved, achieving highly reliable and adaptive identification of leather products.

CN122265812APending Publication Date: 2026-06-23海宁中国皮革城网络科技有限公司
0 Cites 0 Cited by

Patent Information

Application Number
CN202610142330.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-02-02
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Existing image recognition-based leather product identification technologies struggle to achieve accurate identification in complex scenarios. They are affected by external factors such as lighting and coating wear, and their decision-making strategies lack adaptability, making them unsuitable for diverse identification tasks such as authenticity verification and damage assessment.

Method used

The system collects videos of the dynamic surface deformation process of the leather products to be inspected and the reference sample under controlled micro-force, extracts multi-dimensional standardized feature vectors that reflect the inherent physical and mechanical properties of the leather, dynamically generates an adaptive identification model, identifies the product through feature weight vectors and decision thresholds, and receives user-interactive correction instructions to optimize the results.

Benefits of technology

It significantly reduces the misjudgment rate caused by different appearances of the same object, improves the accuracy and reliability of the identification results, and realizes precise services for brand authenticity identification, process traceability and damage assessment. It has high reliability and scenario adaptability.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of image recognition, and particularly discloses a leather product comparison and identification method and system based on image recognition, and the method comprises the steps: collecting a dynamic deformation video of to-be-detected leather and a reference sample under a controlled micro force, extracting a multi-dimensional standardized feature vector reflecting inherent physical and mechanical properties, and carrying out the recognition of the multi-dimensional standardized feature vector; receiving identification task description parameters, mapping core discrimination features from a physical and mechanical feature knowledge base, dynamically instantiating a self-adaptive identification model through a meta-learning model, generating a feature weighting scheme and a dynamic decision threshold, calculating a mechanical feature matching degree, generating a visual preliminary report, and adjusting the scheme in combination with a user interactive correction instruction. And updating a result and feeding back the data optimization meta-model. According to the method, the dimension of the identification basis is raised to the essential mechanical attribute, task adaptive decision and man-machine collaborative optimization are realized, the identification reliability and scene adaptability can be improved, the decision process is transparent and explainable, and the method has various requirements of continuous optimization capability, adaptive authenticity identification, traceability and the like.
Need to check novelty before this filing date? Find Prior Art