|
AUTHOR(S): Jiftah Ben Asher, Arkadi Zilbeman, Benayhu Bar Yosef, Eran Gitler
|
TITLE |
 PDF |
KEYWORDS proximate sensing, Nitrogen percent in leaves, RGB camera |
ABSTRACT The purpose of this paper is to summarize an integrated project that included an algorithm to calculate Nitrogen percent (N%) by image processing, to use it as a tool for a decision support system (DSS) and to load the system on an autonomous robot. The robot moves along crop's rows and performs foliar fertilization. The system presented in this project unifies two very different areas – robotics and agronomy. The robot is controlled by a "computerized manager" that provides several tasks. 1st task is to control the movement of the robot including moving direction (or angle), number of steps between plants in a row and the distance between steps. The 2nd task is to take photos of the crop automatically, by a digital color camera. The camera is mounted on a long rod that carries it above the canopy. Then, via the internet cloud, the robot send the photo to a "server" that is to a located elsewhere for N% analysis. The 3rd task of the robot is to receive the output from the "server" and to perform quantitative foliar fertilization according to the output of the "server" analysis. The agronomic decisions are controlled by the algorithm of the server. The "server" is the "brain of the system". It also has several tasks: 1st task is calculate N% from the photo by its built-in algorithm; 2nd task is to approximate actual N content (kg N ha-1) from the product of N% and potential canopy weight. The potential canopy weight is stored within the "server" as a part of a list of best crop production functions vs. days after planting (DAP) for many different crops. The production functions are based on well documented experiments and are given together with N% vs DAP. The 3rd task is to use the actual nitrogen content and compare it with the potential N contend in order to estimate N deficiency or surplus actual N availability of the crop. The 4th task is to release the calculated results to the "computerized manager" that is expected to response by foliar fertilization at appropriate dose. This unified robotics-agronomy system can grant a good representation and a better understanding of the robot labor in agricultural activities for researchers and engineers from different areas, who could be involved in the design and application of precision agriculture techniques. |
Cite this paper Jiftah Ben Asher, Arkadi Zilbeman, Benayhu Bar Yosef, Eran Gitler. (2016) A Robot and Precision Greenhouse Farming . International Journal of Control Systems and Robotics, 1, 177-180 |
