Reconstrucción 3D de malas hierbas utilizando cámaras de profundidad
The objective of this study was to optimize the positioning angle of a Kinect sensor for reconstructing the three dimensional structure of weeds, using Kinect fusion algorithms to generate a 3D point cloud from the depth video stream. The sensor was mounted in different positions facing the plant in order to obtain depth (RGB-D) images from different angles. The results confirmed the correlation between ground truth (e. g. weed biomass) and the measured area with Kinect. In addition, plant height was accurately estimated with a few centimeters error. However, although the Kinect sensor has shown its ability for plant reconstruction, proper positioning of the sensor is critical for correct reconstruction of plants. The best position of the sensor must be chosen according to the species to be measured and their growth stage. These results suggest that Kinect is a promising tool for a rapid and reliable weed characterization, with several important advantages such as low cost, low power requirement and a high frame rate.
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | comunicación de congreso biblioteca |
| Published: |
XV Congreso de Malherbología
|
| Subjects: | Kinect, Plant structure characterization, Agricultura de precisión, Angle of view, Sensores, |
| Online Access: | http://dx.doi.org/10.13039/501100007273 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|