Plant Localization and Discrimination using 2D+3D Computer Vision for Robotic Intra-row Weed Control

摘要: Abstract. Weed management is vitally important in crop production systems. However, conventional herbicide based weed control can lead to negative environmental impacts. Manual weed laborious and impractical for large scale production. Robotic offers a possibility of controlling weeds precisely, particularly growing near or within rows. A computer vision system was developed based on Kinect V2 sensor, using the fusion two-dimensional textural data three-dimensional spatial recognize localized plants different growth stages. Images were acquired plant species such as broccoli, lettuce and corn at database organize these images. Several feature extraction algorithms which addressed problems canopy occlusion damaged leaves. With our proposed algorithms, features extracted used train background classifiers. Finally, efficiency accuracy classification methods demonstrated validated by experiments.