Method and system for automatic identification and segmentation of zircon-quartz pseudomorphs and quantitative characterization

By combining YOLOv8 and K-Means with ImageJ and Roboflow, the automatic identification and quantitative characterization of zircon-quartz artifacts were achieved, solving the problems of high identification cost and low efficiency in existing technologies and providing a systematic and efficient solution.

CN120564167BActive Publication Date: 2025-10-21CHINA UNIV OF GEOSCIENCES (BEIJING)
0 Cites 0 Cited by

Patent Information

Application Number
CN202510653338.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-05-21
Publication Date
2025-10-21
Estimated Expiration
2045-05-21

AI Technical Summary

Technical Problem

Existing technologies require professional geological experience to identify and quantify zircon-quartz artifacts, are costly, lack a systematic process, and are inefficient.

Method used

The YOLOv8 instance segmentation method was used to establish a model for automatic identification and segmentation of zircon-quartz artifacts, and K-Means unsupervised machine learning was combined for quantitative characterization. ImageJ and Roboflow were used for data cropping and annotation, and deep learning and cluster analysis were used to achieve automated and refined processing.

Benefits of technology

It achieves efficient, low-cost, and systematic automatic identification and quantitative characterization of zircon-quartz artifacts, improves identification efficiency, reduces labor costs, and provides more intuitive microscopic feature analysis.

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

Abstract

The application provides a method and system for automatically identifying and segmenting zircon-quartz pseudomorph and quantitatively characterizing, and belongs to the technical field of mineral identification. Two parts are included: fine identification of zircon-quartz pseudomorph structure, and quantitative characterization of zircon-quartz pseudomorph structure. Through a YOLOv8 instance segmentation method, a model for automatically identifying and segmenting zircon-quartz pseudomorph is established to reconstruct the crystal form boundary of the pseudomorph and obtain the internal mineral area, and on this basis, a K-Means unsupervised machine learning method is provided to quantitatively characterize and analyze the segmented pseudomorph, thereby providing an efficient, low-cost and systematic method for solving the problems of fine identification and quantitative analysis of pseudomorph by combining machine learning with image data related to earth science.
Need to check novelty before this filing date? Find Prior Art