Fruit defect detection method based on image recognition and detection system

Abstract

The invention discloses a fruit defect detection method based on image recognition and a detection system. The system realizes fruit surface defect detection through ultraviolet fluorescence phenomenon excitation, excites ketone substances of defect on the surface of a fruit through ultraviolet light to produce the fluorescence phenomenon, and detects defects which cannot be seen with naked eyes,thereby effectively sorting out high-quality fruits, and meanwhile, the system is provided with a light source, and can perform operation at night, thereby effectively utilizing night time to improvethe operation production efficiency. Image data requiring high operational capability are processed by a PC visual processing terminal through cooperative control of the PC visual processing terminaland a single chip microcomputer control terminal, an underlay with high timeliness requirement is processed by the single chip microcomputer control terminal, and the two terminals realize accuraterecognition and positioning picking on fruits through a wireless communication module, a lamp light structure is simple and convenient to control, and the system has high robustness, accuracy and timeliness, and is simple and practical, and suitable for industrialization promotion.

Description

technical field [0001] The invention relates to the technical field of intelligent fruit picking, in particular to an image recognition-based fruit defect detection method and detection system. Background technique [0002] In the prior art, fruit picking is mainly done manually, so it needs to be operated during the day, and the night time is less utilized. At the same time, the picking workers or corresponding intelligent picking equipment in the prior art only classify and eliminate defects visible to the naked eye, so there are defects such as low sorting efficiency, incomplete fruit classification, and insufficient intelligence of the equipment. Contents of the invention [0003] The main purpose of the present invention is to propose a fruit defect detection method based on image recognition with convenient control and efficient detection. The present invention also proposes a detection system using the fruit defect detection method based on image recognition, which ...

AI Technical Summary

Problems solved by technology

[0002] In the existing technology, fruit picking is mainly done manually, so it needs to be done during the day, and the night time is less used

At the same time, the picking workers or corresponding intelligent picking equipment in the prior art only classify and eliminate defects visible to the naked eye, so there are defects such as low sorting efficiency, incomplete fruit classification, and insufficient intelligence of the equipment.

Images