Promising real-time fruit and vegetable quality detection technologies applicable to manipulator picking process

• Published in: International journal of agricultural and biological engineering Vol. 17; no. 2; pp. 14 - 26
• Main Authors: Jintao, Feng, Qinyi, Yang, Hao, Tian, Zhipeng, Wang, Shijie, Tian, Huirong, Xu
• Format: Journal Article
• Language: English
• Published: Beijing International Journal of Agricultural and Biological Engineering (IJABE) 01.04.2024
• Subjects: Agricultural products; Costs; Deep learning; Food quality; Fruits; Machine learning; Machine vision; Manipulators; Mosaics; Nanomaterials; Nanotechnology; Portable equipment; Production lines; R&D; Real time; Research & development; Sensors; Spectrometers; Spectroscopy; Tactile sensors (robotics); Technology; Trends; Vegetables; Vision systems; Work environment
• DOI: 10.25165/j.ijabe.20241702.7678
• ISSN: 1934-6344

In recent years, worldwide research on fruit and vegetable quality detection technology includes machine vision, spectroscopy, acoustic vibration, tactile sensors, etc. These technologies have also been gradually applied to fruit and vegetable grading and sorting lines in recent years, greatly improving the income of farmers. There have been numerous reviews of these techniques. Most of the published research on fruit and vegetable quality detection technology is still carried out in the laboratory. The emphases have been on quality feature extraction, model establishment and experimental verification. The successful application in the fruit and vegetable sorting production line proves that these studies have high application potential and value, and we look forward to the performance of these sensing technologies in the fruit and vegetable picking field. Therefore, in this paper, based on the future highly automated fruit and vegetable picking mode, we will focus on three kinds of fruit and vegetable quality detection technologies including machine vision, tactile sensor and spectroscopy, to provide some reference for future research. Since there are currently limited cases of detecting quality during the fruit and vegetable picking, experiments performed on prototypes of manipulator, or devices such as Nanocilia sensors, portable spectrometers, etc., which are compact and convenient to mount on manipulator will be reviewed. Several tables and mosaics showing the performance of the three technologies in the detection of fruit and vegetable quality over the past five years have been listed. The performance of each sensing technology was relatively satisfactory in the laboratory in general. However, in the picking scenario, there are still many challenges to be solved. Different from industrial environments, agricultural scenarios are complex and changeable. Fragile and vulnerable agricultural products pose another challenge. The development of portable devices and nanomaterials have become important breakthroughs. Optical and tactile detection methods, as well as the integration of different quality detection methods, are expected to be the trends of research and development.