Abstract
Weeding is important for cultivating crops. However, weeding is a big burden to workers because it is done by hand. Automation of weeding can save labor. In addition, consumers are highly interested in organic vegetables. Cultivating crops without using weedkillers will increase the value of the crops. In this study, we developed a weeding robot which determines weeding areas automatically by identifying the registered plants. To identify the plants, we proposed a method of extracting a single leaf image using multiple lights. The aim of the study is a weeding robot that can be applied to any placement. style of plants, and any type of plants.
