Jui-Feng Yeh;Kuei-Mei Lin;Chen-Yu Lin;Jen-Chun Kang
{"title":"Intelligent Mango Fruit Grade Classification Using AlexNet-SPP With Mask R-CNN-Based Segmentation Algorithm","authors":"Jui-Feng Yeh;Kuei-Mei Lin;Chen-Yu Lin;Jen-Chun Kang","doi":"10.1109/TAFE.2023.3267617","DOIUrl":null,"url":null,"abstract":"In this article, the grades of mangoes were classified using an AlexNet–spatial pyramid pooling network (SPP-Net) with a segmentation algorithm based on a Mask region-based convolutional neural network (R-CNN). Computer vision technologies have begun to be used for fruit grade classification, and this is a major topic of interest in agricultural automation. However, because insufficient fruit grade classification accuracy is achieved with these technologies, manual processing must be performed. The accuracy of fruit grade classification can be enhanced using a Mask R-CNN, SPP-Net, and specific background processing. The designed mango grade classification system contains four modules: 1) a user interface module, 2) an object detection module, 3) an image preprocessing module, and 4) a fruit grade classification module. A camera is used to capture images of mangoes for display on the user interface. The object segmentation module generates a mango shape mask and bounding box by using a Mask R-CNN. The image preprocessing module uses the generated bounding box and mango shape mask to crop the mango and color the background blue. Finally, AlexNet–SPP-Net outputs the fruit grade. We validated the proposed approach by implementing it in mango grade classification and comparing its accuracy with that of relevant existing methods from the literature. According to the experimental results, the proposed approach outperforms the traditional AlexNet-based approach.","PeriodicalId":100637,"journal":{"name":"IEEE Transactions on AgriFood Electronics","volume":"1 1","pages":"41-49"},"PeriodicalIF":0.0000,"publicationDate":"2023-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on AgriFood Electronics","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10119648/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
In this article, the grades of mangoes were classified using an AlexNet–spatial pyramid pooling network (SPP-Net) with a segmentation algorithm based on a Mask region-based convolutional neural network (R-CNN). Computer vision technologies have begun to be used for fruit grade classification, and this is a major topic of interest in agricultural automation. However, because insufficient fruit grade classification accuracy is achieved with these technologies, manual processing must be performed. The accuracy of fruit grade classification can be enhanced using a Mask R-CNN, SPP-Net, and specific background processing. The designed mango grade classification system contains four modules: 1) a user interface module, 2) an object detection module, 3) an image preprocessing module, and 4) a fruit grade classification module. A camera is used to capture images of mangoes for display on the user interface. The object segmentation module generates a mango shape mask and bounding box by using a Mask R-CNN. The image preprocessing module uses the generated bounding box and mango shape mask to crop the mango and color the background blue. Finally, AlexNet–SPP-Net outputs the fruit grade. We validated the proposed approach by implementing it in mango grade classification and comparing its accuracy with that of relevant existing methods from the literature. According to the experimental results, the proposed approach outperforms the traditional AlexNet-based approach.