Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Detection of crop pests and diseases based on deep convolutional neural network and improved algorithm
Wu J., Li B., Wu Z.  ICMLT 2019 (Proceedings of the 2019 4th International Conference on Machine Learning Technologies, Nanchang, China, Jun 21-23, 2019)20-27.2019.Type:Proceedings
Date Reviewed: Sep 29 2020

Especially in large monoculture-based agricultural settings, an outbreak of pests or diseases can have a major impact on yield or quality of a crop. Advances in image processing based on convolutional neural network (CNN) architecture over the past decade have yielded major improvements in the accuracy of image classification, renewing interest in its application to pest and disease detection.

The authors describe the use of CNNs for pest and disease detection, starting with two widely used models: GoogLeNet and AlexNet. Using straightforward approaches with these network types results in classification accuracy of about 80 percent for 38 categories of pests and diseases. Because this is not very high, the authors examine two techniques known for optimizing accuracy: transfer learning and data augmentation.

In transfer learning, the network is trained on a source domain with a large, well-curated set of images. This trained network is then fine-tuned by training it further with images from the target domain.

In data augmentation, new data is generated from the original dataset to increase the size of the training set. The authors select the addition of noise as an augmentation method and explore the performance of the network with different levels of added noise.

This combination of techniques resulted in significantly higher performance, with a detection accuracy of more than 98 percent for a noise standard deviation of three. While the authors’ claim that their method is more efficient and reliable than human classification sounds plausible, no data is provided for the accuracy of human-expert classification.

Reviewer:  Franz Kurfess Review #: CR147071 (2102-0035)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Neural Nets (C.1.3 ... )
 
Would you recommend this review?
yes
no
Other reviews under "Neural Nets": Date
Neural networks: an introduction
Müller B., Reinhardt J., Springer-Verlag New York, Inc., New York, NY, 1990. Type: Book (9780387523804)
May 1 1993
The computing neuron
Durbin R. (ed), Miall C., Mitchison G., Addison-Wesley Longman Publishing Co., Inc., Boston, MA, 1989. Type: Book (9780201183481)
May 1 1993
A practical guide to neural nets
McCord-Nelson M., Illingworth W., Addison-Wesley Longman Publishing Co., Inc., Boston, MA, 1991. Type: Book (9780201523768)
May 1 1993
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy