Diseño de robot cartesiano para procesos de segmentación y fenotipado de plantas con visión artificial.

Abstract

At present, the advances that have been made in the computer vision part and, in turn, machine learning technologies have allowed great scientists who study plants to transform their techniques for obtaining plant phenotyping and to incorporate new ways of obtaining use of computer vision techniques, since as it is one of the most important branches today, it has several functionalities in different sectors, one of them is important for this project, the agricultural sector. The project is based on the search for segmentation algorithms currently using and in turn the design of a Cartesian robot that allows segmentation and phenotyping of plants to be carried out in controlled environments. The project is carried out in order to make detailed processes in individual plants observing different characteristics such as the growth of these plants in different growth conditions, along with this, see the size and evolution of their leaves, their fruits and its floral area, and thus, be able to extract visual characteristics of the plants, which help to quantify or quantify visual traits.