基于机器视觉的水果外观品质检测研究进展

Research progress of fruit appearance quality detection based on machine vision

中文关键词:  机器视觉  水果分级  品质检测

英文关键词:machine vision  fruit grading  quality inspection

中文摘要:

      水果中富含多种营养成分, 随着经济和社会生活水平的提高, 高品质水果越来越受人们的青睐, 其外观品质已经成为影响消费者采购的重要因素。早期我国主要依赖人工对水果进行分级, 效率和准确率较低, 成本和工人劳动强度较大。近年来随着机器视觉技术的不断发展, 大量的学者将视觉技术应用到水果外观品质的检测中, 这种技术具有无损坏、低成本、高效率和操作方便等优点。本文结合国内外学者的研究成果, 梳理了机器视觉在水果外观颜色、形状、大小、缺陷和纹理检测中的应用, 着重介绍了缺陷提取和分类器对水果识别算法的研究进展, 分析了传统视觉分级、机器学习和深度学习的应用特点, 提出了机器视觉技术存在的问题并对未来发展趋势进行了展望, 以期为水果外观品质检测研究提供参考与借鉴。

英文摘要:

      Fruits are rich in various nutrients. With the improvement of economic and social living standards, high-quality fruits are increasingly favored by people, and their appearance quality has become an important factor affecting consumer procurement. In the early stage, China mainly relied on manual grading of fruits, which resulted in low efficiency and accuracy, as well as high costs and labor intensity for workers. In recent years, with the continuous development of machine vision technology, a large number of scholars have applied vision technology to the detection of fruit appearance quality. This technology has advantages such as no damage, low cost, high efficiency, and easy operation. This article combined the research results of domestic and foreign scholars to sort out the application of machine vision in fruit appearance color, shape, size, defects, and texture detection. It focused on the research progress of defect extraction and classifiers on fruit recognition algorithms, analyzed the application characteristics of traditional visual grading, machine learning, and deep learning, proposes the problems of machine vision technology, and looked forward to future development trends, to provide reference and inspiration for research on fruit appearance quality detection.