Abstract
Image identification technology has great significance for forestry production and forestry management. Highly similar object identification tasks, such as tree species with similar leaves, are extremely challenging. Simply using typical Convolutional Neural Networks (CNNs) or simply adding more convolutional layers still performs poorly in the above tasks. In this paper, we present a novel attention mechanism to enhance the CNN for identification of tree species with highly similar leaves. This paper presents a highly discriminative network, namely attention branch based convolutional neural networks (ABCNN), to better distinguish the differences between leaves features. Firstly, we proposed a novel structure, in which an attention branch is added in all block layers of network besides the typical normal branch. Secondly, our attention branch adopts a condensation process to obtain a region of interest (ROI) from global information of input and designs a reconstruction process to amplify the features difference to focus on the ROI. Thirdly, we design a fusion process, which carefully combines the attention branch with a normal branch to improve the network performance in the training process. The proposed ABCNN is tested on special dataset of Leafsnap with highly similar tree leaves. Our approach achieved 91.43% classification accuracy, which is higher than previous methods. Furthermore, ABCNN is also tested on general data set of SVHN and obtains 98.27% classification accuracy, which is the most competitive when considering the lower computational resources for ordinary applications. Both above experiments demonstrate the discrimination and robustness of the proposed method.
Highlights
The use of image processing in forestry production and forestry management can improve the efficiency of forestry works, such as precious tree protection, forest environmental monitoring, forest fire warning, forest pest and disease prevention, forest savings estimation and rational exploit of forest resources [1]–[5]
We report a novel attention mechanism to enhance the state-of-the-art Convolutional Neural Networks (CNNs), which aims to improve the network identification ability for challenging identification tasks, such as the identification of tree species with highly similar leaves
Song et al.: To Identify Tree Species With Highly Similar Leaves in all block layers of network besides the typical normal branch; (2) A condensation process is designed to capture the top information of input and a reconstruction process is designed to amplify the discrimination of highly similar features while keeping the original features; (3) The proposed attention branch can be fused and compatible with state-ofthe-art CNN structure; (4) Experiments on both special data set and general data set demonstrate the discrimination and robustness of the proposed method
Summary
The use of image processing in forestry production and forestry management can improve the efficiency of forestry works, such as precious tree protection, forest environmental monitoring, forest fire warning, forest pest and disease prevention, forest savings estimation and rational exploit of forest resources [1]–[5]. We report a novel attention mechanism to enhance the state-of-the-art CNN, which aims to improve the network identification ability for challenging identification tasks, such as the identification of tree species with highly similar leaves. Y. Song et al.: To Identify Tree Species With Highly Similar Leaves in all block layers of network besides the typical normal branch; (2) A condensation process is designed to capture the top information of input and a reconstruction process is designed to amplify the discrimination of highly similar features while keeping the original features; (3) The proposed attention branch can be fused and compatible with state-ofthe-art CNN structure; (4) Experiments on both special data set and general data set demonstrate the discrimination and robustness of the proposed method.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call