TABLE 1.

Classification of driverless perception and decision system working in orchard.

	Advantage	Shortcoming
GNSS (Guevara et al., 2020; Mao et al., 2022)	It can work in the orchard all day and is completely unaffected by the weather	In the orchard, the loss of signal caused by canopy occlusion, multipath effect, radio frequency interference, etc., results in great errors to GNSS navigation and even led to invalid navigation
Binocular vision (Stefas et al., 2016; Lin et al., 2021; Ma et al., 2021; Vrochidou et al., 2022)	Low cost and abundant information (depth map and RGB map)	The accuracy is poor, and is seriously reduced in dim light and at night, failing to meet the needs of overnight operation in orchards
Lidar (Bergerman et al., 2015; Blok et al., 2019; Jones et al., 2019; Guevara et al., 2020; Zhang et al., 2020)	The cost is high, and is greatly affected by bad weather such as rain and snow	The cost is high, and is greatly affected by bad weather such as rain and snow
Millimeter wave radar (Li X. et al., 2020; Wang et al., 2021)	It has a strong penetrability and is not affected by light, and can meet all kinds of weather in the orchard	The atmospheric attenuation is large and the detection distance is short, so it cannot be perceived in a large range