Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification
Traditional image-centered methods of plant identification could be confused due to various views, uneven illuminations, and growth cycles. To tolerate the significant intraclass variances, the convolutional recurrent neural networks (C-RNNs) are proposed for observation-centered plant identificatio...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2018-01-01
|
| Series: | Journal of Electrical and Computer Engineering |
| Online Access: | http://dx.doi.org/10.1155/2018/9373210 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849683682234728448 |
|---|---|
| author | Xuanxin Liu Fu Xu Yu Sun Haiyan Zhang Zhibo Chen |
| author_facet | Xuanxin Liu Fu Xu Yu Sun Haiyan Zhang Zhibo Chen |
| author_sort | Xuanxin Liu |
| collection | DOAJ |
| description | Traditional image-centered methods of plant identification could be confused due to various views, uneven illuminations, and growth cycles. To tolerate the significant intraclass variances, the convolutional recurrent neural networks (C-RNNs) are proposed for observation-centered plant identification to mimic human behaviors. The C-RNN model is composed of two components: the convolutional neural network (CNN) backbone is used as a feature extractor for images, and the recurrent neural network (RNN) units are built to synthesize multiview features from each image for final prediction. Extensive experiments are conducted to explore the best combination of CNN and RNN. All models are trained end-to-end with 1 to 3 plant images of the same observation by truncated back propagation through time. The experiments demonstrate that the combination of MobileNet and Gated Recurrent Unit (GRU) is the best trade-off of classification accuracy and computational overhead on the Flavia dataset. On the holdout test set, the mean 10-fold accuracy with 1, 2, and 3 input leaves reached 99.53%, 100.00%, and 100.00%, respectively. On the BJFU100 dataset, the C-RNN model achieves the classification rate of 99.65% by two-stage end-to-end training. The observation-centered method based on the C-RNNs shows potential to further improve plant identification accuracy. |
| format | Article |
| id | doaj-art-ced5ad09c2c54086b7f4f8f9fb4ce908 |
| institution | DOAJ |
| issn | 2090-0147 2090-0155 |
| language | English |
| publishDate | 2018-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of Electrical and Computer Engineering |
| spelling | doaj-art-ced5ad09c2c54086b7f4f8f9fb4ce9082025-08-20T03:23:43ZengWileyJournal of Electrical and Computer Engineering2090-01472090-01552018-01-01201810.1155/2018/93732109373210Convolutional Recurrent Neural Networks for Observation-Centered Plant IdentificationXuanxin Liu0Fu Xu1Yu Sun2Haiyan Zhang3Zhibo Chen4School of Information Science and Technology, Beijing Forestry University, Beijing 100083, ChinaSchool of Information Science and Technology, Beijing Forestry University, Beijing 100083, ChinaSchool of Information Science and Technology, Beijing Forestry University, Beijing 100083, ChinaSchool of Information Science and Technology, Beijing Forestry University, Beijing 100083, ChinaSchool of Information Science and Technology, Beijing Forestry University, Beijing 100083, ChinaTraditional image-centered methods of plant identification could be confused due to various views, uneven illuminations, and growth cycles. To tolerate the significant intraclass variances, the convolutional recurrent neural networks (C-RNNs) are proposed for observation-centered plant identification to mimic human behaviors. The C-RNN model is composed of two components: the convolutional neural network (CNN) backbone is used as a feature extractor for images, and the recurrent neural network (RNN) units are built to synthesize multiview features from each image for final prediction. Extensive experiments are conducted to explore the best combination of CNN and RNN. All models are trained end-to-end with 1 to 3 plant images of the same observation by truncated back propagation through time. The experiments demonstrate that the combination of MobileNet and Gated Recurrent Unit (GRU) is the best trade-off of classification accuracy and computational overhead on the Flavia dataset. On the holdout test set, the mean 10-fold accuracy with 1, 2, and 3 input leaves reached 99.53%, 100.00%, and 100.00%, respectively. On the BJFU100 dataset, the C-RNN model achieves the classification rate of 99.65% by two-stage end-to-end training. The observation-centered method based on the C-RNNs shows potential to further improve plant identification accuracy.http://dx.doi.org/10.1155/2018/9373210 |
| spellingShingle | Xuanxin Liu Fu Xu Yu Sun Haiyan Zhang Zhibo Chen Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification Journal of Electrical and Computer Engineering |
| title | Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification |
| title_full | Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification |
| title_fullStr | Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification |
| title_full_unstemmed | Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification |
| title_short | Convolutional Recurrent Neural Networks for Observation-Centered Plant Identification |
| title_sort | convolutional recurrent neural networks for observation centered plant identification |
| url | http://dx.doi.org/10.1155/2018/9373210 |
| work_keys_str_mv | AT xuanxinliu convolutionalrecurrentneuralnetworksforobservationcenteredplantidentification AT fuxu convolutionalrecurrentneuralnetworksforobservationcenteredplantidentification AT yusun convolutionalrecurrentneuralnetworksforobservationcenteredplantidentification AT haiyanzhang convolutionalrecurrentneuralnetworksforobservationcenteredplantidentification AT zhibochen convolutionalrecurrentneuralnetworksforobservationcenteredplantidentification |