A method of detecting apple leaf diseases based on improved convolutional neural network.

Jie Di,Qing Li,Nguyen Quoc Khanh Le

doi:10.1371/journal.pone.0262629

Abstract

Apple tree diseases have perplexed orchard farmers for several years. At present, numerous studies have investigated deep learning for fruit and vegetable crop disease detection. Because of the complexity and variety of apple leaf veins and the difficulty in judging similar diseases, a new target detection model of apple leaf diseases DF-Tiny-YOLO, based on deep learning, is proposed to realize faster and more effective automatic detection of apple leaf diseases. Four common apple leaf diseases, including 1,404 images, were selected for data modeling and method evaluation, and made three main improvements. Feature reuse was combined with the DenseNet densely connected network and further realized to reduce the disappearance of the deep gradient, thus strengthening feature propagation and improving detection accuracy. We introduced Resize and Re-organization (Reorg) and conducted convolution kernel compression to reduce the calculation parameters of the model, improve the operating detection speed, and allow feature stacking to achieve feature fusion. The network terminal uses convolution kernels of 1 × 1, 1 × 1, and 3 × 3, in turn, to realize the dimensionality reduction of features and increase network depth without increasing computational complexity, thus further improving the detection accuracy. The results showed that the mean average precision (mAP) and average intersection over union (IoU) of the DF-Tiny-YOLO model were 99.99% and 90.88%, respectively, and the detection speed reached 280 FPS. Compared with the Tiny-YOLO and YOLOv2 network models, the new method proposed in this paper significantly improves the detection performance. It can also detect apple leaf diseases quickly and effectively.

Highlights

With over 2,000 years of history in China, areas of apple implantation have expanded annually, and 65% of all apples in the world are produced in China
Since target detection technology appeared during the 1960s, target detection based on deep learning has emerged, and convolutional neural network learning methods are being applied to classify and identify apple leaf diseases, such as the SDD model, and the R-CNN and YOLO series of algorithms
The present study improved the TinyYOLO model by combining DenseNet and F-YOLO network concepts and we propose that the DF-Tiny-YOLO apple leaf disease detection model can solve the problem of differentiating apple scab, black rot, and cedar rust from healthy apple leaves

Summary

Introduction

With over 2,000 years of history in China, areas of apple implantation have expanded annually, and 65% of all apples in the world are produced in China. Convolutional neural network learning methods are being applied to classify and identify apple leaf diseases, such as the SDD model, and the R-CNN and YOLO series of algorithms. Based on the DenseNet-121 deep convolutional network, Zhong and Zhao [28] proposed regression, multi-label classification, and focus loss function, to identify apple leaf diseases, with a recognition accuracy rate > 93%, which was better than the traditional multiclassification method based on the cross-entropy loss function. Son [30] proposed a new method based on a region-of-interest-aware deep convolutional neural network (ROI-aware DCNN) to render deep features more discriminative and increase classification performance for apple leaf disease identification, with an average accuracy of 84.3%.

Algorithm principles

DF-Tiny-YOLO network model

Experimental process

Experimental results and analysis

Research conclusion

Future research