An Effective Data Augmentation Strategy for CNN-Based Pest Localization and Recognition in the Field

In agriculture, pest always causes the major damage in fields and results in significant crop yield losses. Currently, manual pest classification and counting are very time-consuming and many subjective factors can affect the population counting accuracy. In addition, the existing pest localization...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access 2019, Vol.7, p.160274-160283
Main Authors: Li, Rui, Jia, Xiufang, Hu, Min, Zhou, Man, Li, Dengshan, Liu, Wancai, Wang, Rujing, Zhang, Jie, Xie, Chengjun, Liu, Liu, Wang, Fangyuan, Chen, Hongbo, Chen, Tianjiao, Hu, Haiying
Format: Article
Language:English
Subjects:
Artificial neural networks
convolutional neural network
Crop yield
Data augmentation
Deep learning
Feature extraction
Image resolution
Insects
Localization
Monitoring
multi-scale
Object detection
Pest localization
pest recognition
Pests
Recognition
Scale models
Strategy
Training
Wheat
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In agriculture, pest always causes the major damage in fields and results in significant crop yield losses. Currently, manual pest classification and counting are very time-consuming and many subjective factors can affect the population counting accuracy. In addition, the existing pest localization and recognition methods based on Convolutional Neural Network (CNN) are not satisfactory for practical pest prevention in fields because of pests' different scales and attitudes. In order to address these problems, an effective data augmentation strategy for CNN-based method is proposed in this paper. In training phase, we adopt data augmentation through rotating images by various degrees followed by cropping into different grids. In this way, we could obtain a large number of extra multi-scale examples that could be adopted to train a multi-scale pest detection model. In terms of test phase, we utilize the test time augmentation (TTA) strategy that separately inferences input images with various resolutions using the trained multi-scale model. Finally, we fuse these detection results from different image scales by non-maximum suppression (NMS) for the final result. Experimental results on wheat sawfly, wheat aphid, wheat mite and rice planthopper in our domain specific dataset, show that our proposed data augmentation strategy achieves the pest detection performance of 81.4% mean Average Precision (mAP), which improves 11.63%, 7.93%,4.73% compared to three state-of-the-art approaches.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2019.2949852