TY - GEN
T1 - A Portable Instrument for Detecting Surface Defects of Agricultural Tools Based on Deep Learning
T2 - 2nd International Conference on Electronic Information Technology and Smart Agriculture, ICEITSA 2022
AU - Huang, Jing
AU - Tan, Yixuan
AU - Hu, Binhan
AU - Chen, Haiting
AU - Yang, Xi
AU - Li, Shu
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Our country is a big agricultural country, the agricultural tools occupy an indispensable position in the process of agricultural production, and the defect detection on the surface of agricultural tools can reduce the impact on agricultural production. We take the aluminum farm tools as an example, in order to realize the quick detection of the surface defects of farm tools and help intelligent agriculture, and reduce the false detection caused by artificial subjective factors, in this paper, a kind of portable instrument for detecting surface defects of farm tools based on depth learning is presented. Taking the lightweight MobileNetV3 as the feature extraction network of the whole model, we aim to design a set of lightweight network model with small model parameters and fast detection time, and deploy the lightweight model on mobile devices, the automatic detection of agricultural tool defects is realized quickly. Abandon the traditional manual testing, to 'Minimize the cost, to ensure the accuracy of detection' as a concept to find a machine-assisted or to replace the possibility of artificial.
AB - Our country is a big agricultural country, the agricultural tools occupy an indispensable position in the process of agricultural production, and the defect detection on the surface of agricultural tools can reduce the impact on agricultural production. We take the aluminum farm tools as an example, in order to realize the quick detection of the surface defects of farm tools and help intelligent agriculture, and reduce the false detection caused by artificial subjective factors, in this paper, a kind of portable instrument for detecting surface defects of farm tools based on depth learning is presented. Taking the lightweight MobileNetV3 as the feature extraction network of the whole model, we aim to design a set of lightweight network model with small model parameters and fast detection time, and deploy the lightweight model on mobile devices, the automatic detection of agricultural tool defects is realized quickly. Abandon the traditional manual testing, to 'Minimize the cost, to ensure the accuracy of detection' as a concept to find a machine-assisted or to replace the possibility of artificial.
KW - Deep Learning
KW - defect detection
KW - farm tools
KW - lightweight
UR - https://www.scopus.com/pages/publications/85149533293
U2 - 10.1109/ICEITSA57468.2022.00039
DO - 10.1109/ICEITSA57468.2022.00039
M3 - 会议稿件
AN - SCOPUS:85149533293
T3 - Proceedings - 2022 2nd International Conference on Electronic Information Technology and Smart Agriculture, ICEITSA 2022
SP - 178
EP - 184
BT - Proceedings - 2022 2nd International Conference on Electronic Information Technology and Smart Agriculture, ICEITSA 2022
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 9 December 2022 through 11 December 2022
ER -
