Evaluation method of rowing performance and its optimization for UAV-based shot seeding device on rice sowing

Abstract

In recent years, unmanned aerial vehicle (UAV) has been widely used in the agricultural production. Combined with the promotion of rice direct seeding technology, UAV rice seeding has become an important planting method. In order to quickly obtain the rowing performance on rice sowing, this study proposes an evaluation method based on image recognition technology. In the process of image recognition, a binary method of dynamic threshold based on mean filter difference (BD-MFD) was adopted, and obtained the position coordinates of the seeds. According to the evaluation method, the key structural parameters and operational parameters of the UAV-based shot seeding device are optimized. In the bench test, the size parameters of guide tube were optimized, and the best seeding effect was obtained when the length and diameter of the guide tube were 100 mm and 17 mm, respectively. Experiments were carried out on the seeding effect under different seed discharge rate levels, and the reasonable range of the seed metering wheel rotational speed was 16–24 rpm. In the field seeding test, the appropriate friction wheel rotation speed at different working heights was obtained. In the two-factor comprehensive test of flying speed and seed metering wheel rotational speed, when the flying speed was 1.0–2.0 m/s and the seed metering rotational wheel speed was 22–26 rpm, it has better seeding uniformity. In the five-row seeding experiment, the average width of seed row (Bd) ranged from 66.17 mm to 75.34 mm, and the C.V. (coefficient of variation) of seeding uniformity was between 18.44 % and 27.04 %. The study provided a new idea for UAV rice seeding, and optimized the structural parameters and operation parameters through experiments.