Pose Correction Engine for Counterfeit Item Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The proliferation of counterfeit limited-release items in the market due to their exclusivity has led to a lack of trust in transactions, as consumers and even sellers often cannot verify the authenticity of these items, hindering industry growth.
Innovation Solution
A Pose Correction Engine that preprocesses user-source images to determine if objects are authentic or counterfeit by generating pose-corrected images, using a combination of segmentation, depth estimation, and registration phases, and training machine learning networks with reference images to align and correct the pose of objects in source images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If users capture source images of physical objects, then authentication verification is enabled, but the images may not be in perfect alignment with predefined poses reducing measurement precision
Solution Approach 1:
The system performs preliminary pose correction by generating a pose corrected image that combines the source image with a reference image. This preliminary alignment process adjusts the source image to match the predefined pose before authentication, ensuring both verification reliability and measurement precision are achieved
Solution Approach 2:
The pose corrected image acts as an intermediary between the user-captured source image and the authentication system. It serves as a mediating representation that has been geometrically transformed to match the expected pose, allowing the authentication system to work with precisely aligned images without requiring users to manually achieve perfect positioning
2Measurement precision
If the system requires perfect alignment with predefined poses, then measurement precision is improved, but ease of operation deteriorates as users must capture images in exact positions
Solution Approach 1:
The system performs self-service pose correction by automatically generating the pose corrected image through combination with a reference image. This eliminates the need for users to manually position objects or adjust camera angles to achieve perfect alignment, as the system autonomously corrects the pose during image processing
Solution Approach 2:
The pose correction is performed as a preliminary processing step before authentication. The system pre-aligns the source image by combining it with a reference image that contains the object in the correct pose, so that subsequent authentication operations work with properly aligned images without requiring user intervention for positioning
Data Source
AI summary
Various embodiments are directed to a Pose Correction Engine (“Engine”). The Engine generates a reference image of the object of interest. The reference image portrays the object of interest oriented according to a first pose. The Engine receives a source image of an instance of the object. The source image portrays the instance of the object oriented according to a variation of the first pose. The Engine determines a difference between the first pose of the reference image and the variation of the first pose of the source image. The Engine identifies, based on the determined difference, one or portions of a three-dimensional (3D) map of a shape of the object obscured by the variation of the first pose portrayed in the source image. The Engine generates a pose corrected image of the instance of the object that portrays at least a portion of the source image and at least the identified portion of the 3D map of the shape of the object.