A method and device for online detection of potato quality based on three-dimensional reconstruction

By using 3D reconstruction technology, a 3D model of potatoes is created through camera shooting and image processing, which solves the problems of low efficiency and insufficient accuracy in existing detection methods and realizes efficient and accurate potato quality grading.

CN120746982BActive Publication Date: 2026-07-14HUAZHONG AGRI UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUAZHONG AGRI UNIV
Filing Date
2025-06-23
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing potato quality testing methods are inefficient, have low classification accuracy, and involve complex calculations, making it impossible to achieve efficient and precise quality grading.

Method used

An online potato quality detection method based on 3D reconstruction is adopted. The video is captured by a camera, the potato axial rotation image is segmented and processed, and then the image is processed and the color threshold is extracted. A Cartesian coordinate system is established, a 3D model is reconstructed, and the quality is calculated by combining the density formula to achieve accurate grading.

Benefits of technology

It enables efficient and accurate potato quality testing and grading, reduces manpower consumption, minimizes testing damage, and improves testing accuracy and efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120746982B_ABST
    Figure CN120746982B_ABST
Patent Text Reader

Abstract

The application discloses a kind of based on three-dimensional reconstruction's potato quality on-line detection method and device, it is related to agricultural machinery technical field, the on-line detection method includes: through camera shooting potato video;The segmented processing of the potato video of shooting;From the potato video after segmentation processing with equal time interval, corresponding quantity of potato image is extracted, and from the image after processing, according to the color threshold of potato area, potato mask image is extracted, and the potato image after removing background is obtained;The angle of potato rotation axis deflection is calculated, and the potato image is corrected;With the corrected potato image, Cartesian coordinate system is established, and potato contour coordinate point is extracted;Three-dimensional profile is reconstructed using Poisson reconstruction method, and the three-dimensional model of potato is generated;The quality of potato test sample is divided;The volume of potato three-dimensional model is calculated;According to density formula, the three-dimensional reconstruction quality of potato is calculated, based on the potato volume model established, the grading test of potato test sample is carried out.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • On-line rapid measurement method for volume of potato

    CN116309787A

  • Deep-learning-based SLM online quality monitoring and repairing methods and deep-learning-based SLM online quality monitoring and repairing system

    US20240370002A1