基于改进Faster R-CNN模型的丁岙杨梅成熟度检测方法  

Improved Faster R-CNN model-based maturity detection method for Ding’ao bayberry

作　　者：刘玉耀 彭琼尹 LIU Yu-yao;PENG Qiong-yin(College of Artificial Intelligence,Zhejiang Dongfang Polytechnic,Wenzhou 325000,Zhejiang,China)

机构地区：[1]浙江东方职业技术学院人工智能学院,浙江温州325000

出　　处：《湖北农业科学》2025年第4期7-13,30,共8页Hubei Agricultural Sciences

基　　金：浙江省教育厅一般科研项目基金资助项目(Y202352993)。

摘　　要：为了在复杂的自然生长环境中快速、精准地实现丁岙杨梅(Myrica rubra)不同成熟度检测,提出基于改进Faster R-CNN模型(ConvNeXt-T+SE+FPN)的丁岙杨梅成熟度检测方法。采用ConvNeXt-T作为主干特征提取网络,提升复杂场景下的检测能力;引入了SE注意力机制和特征金字塔网络(FPN),增强模型对丁岙杨梅不同成熟度特征敏感性以及小目标果实的检测能力。相较于ResNet50,ConvNeXt-T+SE、ConvNeXt-T+FPN、ConvNeXt-T+SE+FPN能够使模型的平均精度均值(mAP)分别提升14.75%、19.85%、21.86%,其中ConvNeXt-T+SE+FPN的mAP提升幅度最大,能够有效提高丁岙杨梅不同成熟度的检测性能。通过对丁岙杨梅图像数据集进行训练和测试,改进Faster R-CNN模型在不同成熟度果实的检测中表现出较高的准确性,对未成熟、半成熟、近成熟和全成熟果实识别的平均精度(AP)分别为96.90%、94.63%、95.91%、97.58%,mAP为96.26%;相比Faster R-CNN模型,改进Faster R-CNN模型的mAP提升了21.86%。改进Faster R-CNN模型能够有效提升丁岙杨梅成熟度的检测精度,给杨梅的智能化采摘提供有力支持。To rapidly and accurately detect the maturity levels of Ding’ao bayberry(Myrica rubra)in complex natural growth environments,an improved Faster R-CNN model(ConvNeXt-T+SE+FPN)-based maturity detection method was proposed.ConvNeXt-T was adopted as the backbone feature extraction network to enhance detection capabilities in complex scenarios.The SE attention mechanism and Feature Pyramid Network(FPN)were introduced to improve the model’s sensitivity to maturity-related features and detection of small-target fruits.Compared to ResNet50,ConvNeXt-T+SE,ConvNeXt-T+FPN,and ConvNeXt-T+SE+FPN increased the mean average precision(mAP)by 14.75%,19.85%,and 21.86%,respectively.The ConvNeXt-T+SE+FPN configuration achieved the largest mAP improvement,effectively enhancing detection performance for different maturity levels of Ding’ao bayberry.Through training and testing on the Ding’ao bayberry image dataset,the improved Faster R-CNN model demonstrated high accuracy in detecting different maturity levels.The average precision(AP)for unripe,semi-ripe,near-ripe,and fully ripe fruit recognition was 96.90%,94.63%,95.91%,and 97.58%,respectively,with an mAP of 96.26%.Compared to the original Faster R-CNN model,the improved model achieved a 21.86%increase in mAP.The improved Faster R-CNN model effectively enhanced the detection accuracy of Ding’ao bayberry maturity,providing strong support for intelligent harvesting of bayberry fruits.

关 键 词：丁岙杨梅(Myrica rubra) 改进Faster R-CNN模型 成熟度 检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]