A farmer-assistant robot for nitrogen fertilizing management of greenhouse crops

作者：Vakilian Keyvan Asefpour; Massah Jafar^*

来源：Computers and Electronics in Agriculture, 2017, 139: 153-163.

DOI：10.1016/j.compag.2017.05.012

• Image processing
• Cucumber
• Robotic system
• Nitrogen use efficiency
• Image textural features

In this study, an autonomous machine vision-based system was developed for precise nitrogen fertilizing management to improve nitrogen use efficiency in greenhouse crops. Two scenarios were considered: in the first scenario, nitrogen fertilization was done based on available instructions and protocols provided by the seed producer. The second scenario was nitrogen fertilization using a machine vision-based robot which was moving between the crop rows to monitor plants' needs and requirements. The scenarios were examined in a hydroponic greenhouse for four varieties of cucumber. The extracted plant image features, namely entropy, energy, and local homogeneity were the markers for precise timing of nitrogen fertilization in the cucumber crops. In scenario 2, nitrogen fertilizing was based on a signal transmitted from the robot to a wireless receiver when at least one of the mean values of normalized image textural features of cucumber crops had a difference more than a (a = 10, 15, and 20%) with the same feature in scenario 1. Experimental results showed that scenario 2 for a = 15% decreased the nitrogen fertilizer consumption about 18% without lowering the fruit yield or fruit quality parameters including firmness, total soluble solids, chlorophyll and ascorbic acid contents. This scenario can be considered as an efficient nitrogen fertilizing management method in commercial greenhouses using a low-cost machine vision based robotic system.

• 出版日期2017-6-15