Non-destructive fruit defect detection method and system based on neural networks

US12169190B1 · kind B1 · utility

Assignee

• INSTITUTE OF FACILITY AGRICULTURE, GUANGDONG ACADEMY OF AGRICULTURAL SCIENCES · CN

Inventors

• Sai Xu · Nanjing, CN
• Huazhong Lu · 流花街道, CN
• Xin Liang · Beijing, CN

Key dates

• Filing date: Jul 4, 2024
• Grant date: Dec 17, 2024
• Expiry date: Jul 4, 2044

Classification

• Technology area (CPC Y): Emerging Cross-Sectional Technologies
• CPC primary: Y02P90/30
• WIPO field: Measurement
• WIPO sector: Instruments

Abstract

A non-destructive fruit defect detection method and system based on neural networks are used to solve problem of inaccurate selection of high-quality fruits by current consumers. The system includes a standard formulation module configured to formulate monitoring standards for different batches and varieties of the fruits to obtain standard detection parameters for the different batches and varieties of the fruits, a preliminary identification module configured to preliminarily identify external conditions of the different batches and varieties of the fruits, a non-destructive detection module configured to non-destructively detect the different batches and varieties of the fruits, generate a fruit abnormal signal or obtain growth deviation values of the different batches and varieties of the fruits, and a quality judgment module configured to judge quality of the different batches and varieties of the fruits. Accurate non-destructive detection for the different batches and varieties of the fruits are realized.