摘要
针对农贸市场、果蔬超市中结算流程不够智能化以及重型神经网络模型部署困难等问题,对果蔬分类模型轻量化识别方法进行了研究。首先针对果蔬智能识别设备所在环境差异大、果蔬套袋问题,采用多场景采集方案在果蔬超市现场采集果蔬170种、图片136 000张,并设计了弱化套袋的图像预处理方法,对数据进一步增强。然后针对重量级神经网络部署困难以及成本较高的问题,设计了一种基于深度可分离卷积的果蔬分类识别模型,并进行训练测试,其Top-1准确率达96.8%,Top-5准确率达100%,相对于Mobilenetv2-224,运算量减少了70%,相对于Mobilenetv3-224,运算量减少了60%,识别能力介于Mobilenetv2-224和Mobilenetv3-224之间。最后对所设计果蔬分类模型在实际部署中面临的问题进行了分析。
Aiming at the problem that the settlement process in agricultural trade markets and fruit and vegetable supermarkets was not intelligent enough and the difficulty of deploying heavy neural network models,the lightweight recognition method of fruit and vege⁃table classification model was studied.Firstly,in response to the large differences in the environment where the intelligent recognition equipment for fruits and vegetables was located,and the problem of fuzzy features in fruit and vegetable bagging,a multi-scene collec⁃tion scheme was used to collect 170 kinds of fruits and vegetables and 136000 pictures in the fruit and vegetable supermarket,and an image preprocessing scheme for weakened bagging was formulated to further enhance the data.Secondly,aiming at the difficulty of de⁃ploying the heavyweight neural network and the high cost,a fruit and vegetable classification recognition model based on Depthwise separable convolution was designed,trained and tested.Its Top-1 success rate had reached 96.8%,and the Top-5 success rate had reached 100%.Compared to Mobilenetv2-224,the amount of computation had been reduced by 70%,compared to Mobilenetv3-224,the amount of computing had also been reduced by 60%,and the recognition ability was higher than Mobilenetv2-224 and lower than Mobilenetv3-224.Finally,the problems faced by the designed fruit and vegetable classification model in the actual deployment were analyzed.
作者
岳振
李卓然
王绪谦
侯宗升
苗壮
郑毅
刘杰
YUE Zhen;LI Zhuo-ran;WANG Xu-qian;HOU Zong-sheng;MIAO Zhuang;ZHENG Yi;LIU Jie(School of Science and Information Science,Qingdao Agricultural University,Qingdao 266109,Shandong,China;Qingdao Topscomm Communication Co.,Ltd.,Qingdao 266109,Shandong,China;Yiwen Family Farm in Pingyi County,Linyi 273302,Shandong,China)
出处
《湖北农业科学》
2024年第8期28-34,53,共8页
Hubei Agricultural Sciences
基金
山东省自然科学基金面上项目“基于图自注意网络的图嵌入聚类技术研究”(ZR2021MF078)
青岛农业大学博士课题“农产品分选中的轻量级深度学习模型研究”。
关键词
果蔬分类
图像增强
深度可分类卷积
轻量化神经网络
fruit and vegetable classification
image enhancement
depthwise separable convolution
lightweight neural network
