Development and evaluation of a seed position mapping system

Abstract

With the development of precision agriculture, individual plant care systems have become an important research topic. These systems require accurate and reliable information about the position of individual plants. Seed position mapping, an important method of accurate plant positioning techniques, can obtain seed location data directly in planting stage. The utilization of seed location data is maximized by providing references for the field managements such as intra-row weeding, fertilizer and pesticide application. This study presented the design of a seed mapping system for corn planters in high-position seeding method (HSM). An experimental study of the designed mapping system was performed to evaluate its performance with a four-row planter. The seed firmer (SF) which can reduce seed bouncing and the limiting plate (LP) which can disable the parallel four-bar linkage (PFL) are used to modify the four-row planter. Each planter unit had been modified differently (with SF and LP added, without SF and LP added, with only LP added, with only SF added). The accuracy of the seed position mapping system was evaluated for different operation speed (3, 5, and 7 km/h), seed spacing (20, 30, and 40 cm), and the global positing system (GPS) frequency (1, 5, and 10 Hz) by the distance deviation (DD) and the location quality index (LQI), where DD is the distance deviation between the predicted and actual seed locations, while LQI is defined as the percentage of seeds with a DD < 5 cm. Field tests show that the application of SF can decrease DD and increase LQI immensely: the average decrease of DD is 19.85 mm in the row with SF compared with that without SF, while the average increase of LQI is 29.4%. The application of LP has no significant impact on DD and LQI under experimental conditions. The DD was highly significant influenced by operation speed and seed spacing (p < 0.01), but significant influenced by GPS frequency (p < 0.05). The LQI was highly significant influenced by operation speed (p < 0.01), significant influenced by seed spacing (p < 0.05), but not significant influenced by GPS frequency (p > 0.05). Slower operation speed and larger seed spacing help to improve the mapping accuracy, achieving lower DD and higher LQI. The maximum operating speed of proposed mapping system is 3 km/h, the average DD and LQI in the presence of SF under this condition were 24.4 mm and 91.3%. The mapping accuracy under such conditions is satisfactory.