Open AccessResearch on Intelligent recognition system of Cotton apical Bud based on Deep Learning
Author(s) -
Jianliang Li,
Xinlei Zhi,
Yingying Wang,
Qingzheng Cao
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1820/1/012134
Subject(s) - topping , automation , computer science , identification (biology) , artificial intelligence , engineering , horticulture , botany , biology , mechanical engineering
Cotton topping is a key link in the whole cotton production process, which can eliminate the top growth advantage and promote cotton production and income. The automation of cotton topping can greatly reduce the topping time and labor intensity, and the high-speed and accurate identification of cotton top buds is the prerequisite and basis for automatic topping. This article designs a cotton top bud intelligent identification system based on YOLOv3 network, which can detect cotton top buds in visible light images in real time, and provide visual information for the subsequent realization of cotton top bud position measurement and mechanical control. A computer workstation (RTX2070s) was used to identify 15049 cotton top bud images under different weather and illumination conditions. The results show that the average time of single frame image recognition is controlled within 100 milliseconds, and the top bud recognition rate reaches 96%, which creates good conditions for the development of automatic cotton topping equipment and has broad application prospects.