DDLA: dual deep learning architecture for classification of plant species
DDLA: dual deep learning architecture for classification of plant species
复制标题
DOI：
10.1049/iet-ipr.2019.0346
复制
发表时间：
2019-10-17
影响因子：
2.3
通讯作者：
Vajravelu, Sathiesh Kumar
中科院分区：
文献类型：
--
作者：
Sundara Sobitha Raj, Anubha Pearline;Vajravelu, Sathiesh Kumar
植物物种识别使用双重深度学习架构（DDLA）方法进行。DDLA由MobileNet和DenseNet-121架构组成。从各个体系结构中获得的特征向量被连接起来形成最终的特征向量。然后使用机器学习（ML）分类器（如线性判别分析、多项逻辑回归（LR）、朴素贝叶斯、分类和回归树、k近邻、随机森林分类器、袋化分类器和多层感知器）对提取的特征进行分类。研究中考虑的数据集是标准的（Flavia， Folio和Swedish Leaf）和定制收集的（Leaf-12）数据集。MobileNet和DenseNet-121架构也被用作特征提取器和分类器。结果表明，采用LR分类器的DDLA结构对Flavia、Folio、Swedish leaf和leaf -12数据集的准确率分别为98.71、96.38、99.41和99.39%。与其他方法（DDLA + ML分类器、MobileNet + ML分类器、DenseNet-121 + ML分类器、MobileNet +全连接层（FCL）、DenseNet-121 + FCL）相比，DDLA + LR的准确率更高。研究还发现，与其他方法相比，采用LR分类器的DDLA体系结构在相当的计算时间内获得了更高的精度。
Plant species recognition is performed using a dual deep learning architecture (DDLA) approach. DDLA consists of MobileNet and DenseNet-121 architectures. The feature vectors obtained from individual architectures are concatenated to form a final feature vector. The extracted features are then classified using machine learning (ML) classifiers such as linear discriminant analysis, multinomial logistic regression (LR), Naive Bayes, classification and regression tree, k-nearest neighbour, random forest classifier, bagging classifier and multi-layer perceptron. The dataset considered in the studies is standard (Flavia, Folio, and Swedish Leaf) and custom collected (Leaf-12) dataset. The MobileNet and DenseNet-121 architectures are also used as a feature extractor and a classifier. It is observed that the DDLA architecture with LR classifier produced the highest accuracies of 98.71, 96.38, 99.41, and 99.39% for Flavia, Folio, Swedish leaf, and Leaf-12 datasets. The observed accuracy for DDLA + LR is higher compared with other approaches (DDLA + ML classifiers, MobileNet + ML classifiers, DenseNet-121 + ML classifiers, MobileNet + fully connected layer (FCL), DenseNet-121 + FCL). It is also observed that the DDLA architecture with LR classifier achieves higher accuracy in comparable computation time with other approaches.