Open AccessEvaluation of the accuracy of the recognition algorithm of an autonomous robotic device to control weeds
Author(s) -
A. I. Dyshakov,
M. A. Mirzaev
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1129/1/012056
Subject(s) - computer science , segmentation , weed control , artificial intelligence , weed , control (management) , pattern recognition (psychology) , computer vision , agronomy , biology
The main problem in the agro-industrial sector is the use of harmful herbicides, resulting in some of the weeds becoming more resistant to common herbicides and therefore using more and more powerful and expensive options. To overcome this problem, t is necessary to reduce the amount of herbicide use. The proposed study aims to develop a system that can detect and map weeds and perform mechanically applied spot spraying. The weed detection techniques described in this study include several aspects: imaging, a new algorithm for detecting plants through contour and colour segmentation. The method developed can detect weeds at early stages of growth. Further development of the methods is aimed at solving problems related to complex background without soil treatment, faster image acquisition speed and shorter processing time for spot injection in real time. The results of the experiment show that the proposed hybrid method can be used as a tool for weed control.