| 基于全卷积神经网络的荔枝表皮缺陷提取 |
| |
| 引用本文： | 王佳盛,陈燕,曾泽钦,李嘉威,刘威威,邹湘军.基于全卷积神经网络的荔枝表皮缺陷提取[J].华南农业大学学报,2018,39(6):104-110. |
| |
| 作者姓名： | 王佳盛 陈燕 曾泽钦 李嘉威 刘威威 邹湘军 |
| |
| 作者单位： | 华南农业大学工程学院/南方农业机械与装备关键技术教育部重点实验室 |
| |
| 基金项目： | 国家重点研发计划（2018YFD0101001）；国家自然科学基金（31571568）；广东省科技计划（2015A020209120，2015A020209111） |
| |
| 摘 要： | 【目的】增强荔枝表皮缺陷提取效果,满足其品质检测分级准确性要求。【方法】采用Tensorflow框架构建基于AlexNet的全卷积神经网络AlexNet-FCN,以ReLU为激活函数,Max-pooling为下采样方法,Softmax回归分类器的损失函数作为优化目标,建立荔枝表皮缺陷提取的全卷积神经网络模型,并用批量随机梯度下降法对模型进行优化。【结果】模型收敛后在验证集上裂果交并比(IoUd)为0.83,褐变交并比(IoUb)为0.60,褐变与裂果的总体交并比(IoUa)为0.68;与利用线性SVM、朴素贝叶斯分类器缺陷提取效果相比,该模型的特征提取能力显著提高。【结论】全卷积神经网络在水果表面缺陷提取中具有良好的应用前景。
|
| 关 键 词： | 荔枝 图像处理 缺陷提取 深度学习 全卷积神经网络 品质检测 |
| 收稿时间： | 2018/3/20 0:00:00 |
Extraction of litchi fruit pericarp defect based on a fully convolutional neural network |
| |
| WANG Jiasheng,CHEN Yan,ZENG Zeqin,LI Jiawei,LIU Weiwei and ZOU Xiangjun.Extraction of litchi fruit pericarp defect based on a fully convolutional neural network[J].Journal of South China Agricultural University,2018,39(6):104-110. |
| |
| Authors: | WANG Jiasheng CHEN Yan ZENG Zeqin LI Jiawei LIU Weiwei and ZOU Xiangjun |
| |
| Institution: | College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China,College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China,College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China,College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China,College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China and College of Engineering, South China Agricultural University/Key Laboratory of Key Technology on Agricultural Machine and Equipment, Ministry of Education, Guangzhou 510642, China |
| |
| Abstract: | |
| |
| Keywords: | litchi image processing defect extraction deep learning fully convolutional neural network quality detection |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《华南农业大学学报》浏览原始摘要信息 |
| 点击此处可从《华南农业大学学报》下载免费的PDF全文 |
|
