Luo Dan, Luo Rong, Cheng Jie, Liu Xin
College of Intelligence and Computing, Tianjin Ren'ai University, Tianjin, Tianjin, China.
School of Intelligent Computing Engineering, Changji University, Changji, Xinjian, China.
Front Plant Sci. 2024 Sep 20;15:1415095. doi: 10.3389/fpls.2024.1415095. eCollection 2024.
The grading of products is important in many ways. One of the important activities after harvesting agricultural products is product grading based on shape and color dimensions. This activity in the agricultural transformation industries, Bas Controller, improves various processes on fruits and vegetables with the same dimensions, which improves the storage conditions of the product, creates added value for the farmer, and gives the consumer the power to choose. The main focus of this study is the application of image processing in the field of identification and classification of fruits. It is an application that has received much less attention than other applications of image processing. The proposed systems presented in this article, are software solutions based on image processing techniques, including histogram matching techniques, for detection, Sable edge detection algorithms, Private edge and Kenny edge, Otsu threshold limit, and clustering method It is an automatic mean and classification of different degrees of fruit. In addition, it has been mentioned more about the examination and description of product grading and clustering methods, that by using the proposed application hardware and its connection with the software, a big step can be taken in product quality grading. This method can be used in product classification and packaging. The accuracy rate for peaches, lemons, apples, and tomatoes is 94.58%, 88.23%, 70%, and 93.33%, respectively. The best accuracy for all 20 sample levels is for peach fruit.
产品分级在很多方面都很重要。收获农产品后的一项重要活动是基于形状和颜色维度进行产品分级。在农业加工行业“巴斯控制器”中，这项活动改善了尺寸相同的水果和蔬菜的各种加工流程，改善了产品的储存条件，为农民创造了附加值，并赋予消费者选择权。本研究的主要重点是图像处理在水果识别和分类领域的应用。这是一个比图像处理的其他应用受到少得多关注的应用。本文提出的系统是基于图像处理技术的软件解决方案，包括直方图匹配技术、用于检测的沙布尔边缘检测算法、私有边缘和肯尼边缘、大津阈值极限以及聚类方法，它能自动对不同程度的水果进行均值和分类。此外，更多地提到了产品分级和聚类方法的检验与描述，即通过使用所提出的应用硬件及其与软件的连接，在产品质量分级方面可以迈出一大步。这种方法可用于产品分类和包装。桃子、柠檬、苹果和西红柿的准确率分别为94.58%、88.23%、70%和93.33%。所有20个样本水平中，桃子的准确率最高。
