Deep learning-based automatic recognition network of agricultural machinery images

被引:24
作者
Zhang, Ziqiang [1 ,2 ,3 ]
Liu, Hui [1 ]
Meng, Zhijun [2 ]
Chen, Jingping [2 ]
机构
[1] Capital Normal Univ, Informat Engn Coll, Beijing 100048, Peoples R China
[2] Natl Engn Res Ctr Informat Technol Agr, Beijing 100097, Peoples R China
[3] Beijing Nongke Mansion Room B-508, Beijing 100097, Peoples R China
基金
中国国家自然科学基金;
关键词
Agricultural machinery images; Deep learning; Image recognition; Inception_v3 network; AMTNet; NEURAL-NETWORK;
D O I
10.1016/j.compag.2019.104978
中图分类号
S [农业科学];
学科分类号
09 ;
摘要
Due to the massive amount of data generated by the mobile Internet and the development of large-scale computing devices and technologies, the deep learning algorithm has experienced a breakthrough in terms of image recognition technology. Traditional image recognition requires the complex extraction of image features, whereas deep learning technology can automatically learn image features through multi-layer nonlinear transformation, which is especially proficient at extracting complex global features. An image annotation dataset containing the images of seven types of machines and six types of abnormal images was constructed in this study from the large number of machine images in the agricultural machinery operation supervisory service system. To improve the Inception_v3 network, a network called AMTNet was designed and trained for automatic recognition of agricultural machinery images. Under the same experimental conditions, AMTNet achieved recognition accuracies of 97.83% and 100% on validation sets Top _1 and Top_5, respectively, demonstrating better performance than the classic networks ResNet_50 and Inception v3. To further test the performance of AMTNet, 200 images of each of the 13 types of machine images were selected as test sets. The average area under the curve and Fl-score of the network for image recognition of various machines reached 92% and 96%, respectively. According to the test results, AMTNet shows good robustness to illumination, environmental changes, and small area occlusion, which meets the practical application requirements of intelligent supervision over agricultural machinery operation.
引用
收藏
页数:11
相关论文
共 34 条
[1]  
[Anonymous], 2015, ICLR
[2]  
[Anonymous], NOT PUBL 13 5 YEAR P
[3]  
[Anonymous], 2003, Image and vision computing New Zealand
[4]  
[Anonymous], JIANGSU AGR SCI
[5]  
[Anonymous], 2016, CVPR
[6]  
[Anonymous], 2015, DEEP LEARNING NATURE, DOI [10.1038/nature14539, DOI 10.1038/NATURE14539]
[7]  
[Anonymous], J AGR MECH RES
[8]  
[Anonymous], ADV NEURAL INFORM PR
[9]  
[Anonymous], COMPUTER VISION PROB
[10]  
[Anonymous], 2014, P 2 INT C LEARNING R