Implementación de una estación prototipo con visión artificial, aplicado a la agricultura de precisión

Abstract

Precision agriculture is an activity that, through the observation of environmental variables, allows precise and timely action on agricultural plots in order to achieve greater and more efficient production. In Ecuador, this is an activity little exploited by small and medium producers, both for lack of knowledge of the technologies that can be used for this purpose and for lack of the necessary economic resources to carry it out. Although there are several technologies that are helping to deploy precision agriculture, in the present thesis work, emphasis will be placed on three of them. First, image processing techniques are used for the detection of fruits of a plot. Subsequently, two technologies are used that will take this project to a current field of research, such as the development of applications based on the Internet of Things (IoT). Specifically, cloud computing for the management of information generated by sensors and the Message Queue Telemetry Transport (MQTT) protocol for data messaging. Thus, the design and construction of a prototype station is proposed, with the ability to monitor environmental variables, develop video surveillance tasks as well as capture and process images for the detection of fruits. This station will send the data collected through the Internet to a web application hosted in the IBM Cloud; finally, the results can be monitored by a user through the browser.