Design and Experiment of an Automatic Row-Oriented Spraying System Based on Machine Vision for Early-Stage Maize Corps

Spraying pesticides using row alignment in the maize seedling stage can effectively improve pesticide utilization and protect ecological environment. Therefore, this study extracts a guidance line for crops machine vision develops an automatic row-oriented control system based on high-clearance sprayer. First, feature points of crop rows are extracted vertical projection method. Second, candidate obtained Hough transform, two auxiliary extraction methods slope outlier algorithm proposed. Then, is fitted tangent formula. To greatly robustness algorithm, Kalman filter used to estimate optimize obtain parameters. Finally, visual spraying platform autonomous navigation built, accuracy performance tested. The experimental results showed that, when turned on, average time consumption detection 42 ms, optimal recognition 93.3%, deviation error simulated 3.2 cm that field 4.36 cm. test meet requirements system, it was found decreased with increasing vehicle speed. innovative spray compared traditional spray, inter-row savings were 20.4% 11.4% overall, application significantly improved.