Identifying plant diseases using deep transfer learning and enhanced lightweight network
- PDF / 2,208,054 Bytes
- 19 Pages / 439.37 x 666.142 pts Page_size
- 63 Downloads / 220 Views
Identifying plant diseases using deep transfer learning and enhanced lightweight network Junde Chen 1
1
& Defu Zhang & Y. A. Nanehkaran
1
Received: 7 April 2020 / Revised: 14 July 2020 / Accepted: 18 August 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Plant diseases can cause significant reductions in both the quality and quantity of agricultural products, and they have a disastrous impact on the safety of food production. In severe cases, plant diseases may even lead to no grain harvest completely. Therefore, seeking fast, automatic, less expensive and accurate methods to detect plant diseases is of great realistic significance. In this paper, we studied the transfer learning for the deep CNNs and modified the network structure to enhance the learning ability of the tiny lesion symptoms. The pretrained MobileNet-V2 was extended with the classification activation map (CAM), which was used for visualization as well as plant lesion positioning, and both were selected in our approach. Particularly, the transfer learning was performed twice in model training: the first phase only inferred the weights from scratch for new extended layers while the bottom convolution layers were frozen with the parameters trained from ImageNet; the second phase retrained the weights using the target dataset by loading the model trained in the first phase. Then, the yielded optimum model was used for identifying plant diseases. Experimental results demonstrate the validity of the proposed approach. It achieves an average recognition accuracy of 99.85% on the public dataset. Even under multiple classes and complex background conditions, the average accuracy reaches 99.11% on the collected plant disease images. Thus, the proposed approach efficiently accomplished plant disease identification and presented a superior performance relative to other state-of-the-art methods. Keywords Plant disease identification . Transfer learning . Convolutional neural networks . Image classification
* Defu Zhang [email protected]
1
School of Informatics, Xiamen University, Xiamen 361005, China
Multimedia Tools and Applications
1 Introduction Plant diseases have a devastating effect on agricultural products, and if they are not detected in time, there will be an increase in food insecurity [11]. Especially, the main food crops, including rice and maize, are very important for guaranteeing the food supply and agricultural production. Apart from being a source of carbohydrates for humans, the main crops are also planted for animal food, cooking oil, industrial raw materials and so on [27]. Whereas, crops are quite susceptible by diverse diseases as well. Serious plant diseases have a disastrous impact on the plants and can cause grain failure entirely. The early warning or forecast is the basis of effective prevention and control for the plant diseases; it can reduce the cost of plant diseases and avoid unnecessary pesticide use. Therefore, obtaining information about real-time plant diseases is highlighted in
Data Loading...
