AUTHOR=Hu Xiaobo , Wang Rujing , Du Jianming , Hu Yimin , Jiao Lin , Xu Taosheng 

TITLE=Class-attention-based lesion proposal convolutional neural network for strawberry diseases identification

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2023.1091600

DOI=10.3389/fpls.2023.1091600

ISSN=1664-462X

ABSTRACT=Identifying diseases of strawberries in field is challenging due to the complex background interference and subtle inter-class differences. A feasible method is to segment strawberry lesions from the background and learn more fine-grained features of the lesions. Following this idea, we present a novel Class-Attention-based Lesion Proposal Convolutional Neural Network (CALP-CNN), which utilizes a class response map to locate the main lesion object and propose discriminative lesion details. Specifically, the CALP-CNN firstly locates the main lesion object from the complex background through a class object location module (COLM) and then applies a lesion part proposal module (LPPM) to propose the discriminative lesion details on the located main lesion object. The segmentation and zooming-in operations are adopted between the modules to make the lesion features more significant and benefit to similar disease identification. With a cascade architecture, the CALP-CNN can simultaneously address the interference from the complex background and the misclassification of similar diseases. A series of experiments on a self-built dataset of field strawberry diseases demonstrate that CALP-CNN has better performance than other state-of-the-art attention-based fine-grained image recognition methods. The classification results show that the CALP-CNN can achieve an overall F1-score of 90.86% (6.48% higher than the sub-optimal baseline MMAL-Net), suggesting that the proposed methods are effective in identifying strawberry diseases in the field.