Monitoring and automation modules for a hydroponic system using fuzzy logic and computer vision

Abstract

Hydroponic systems are related to the cultivation of plants without the need for soil, simply complying with the techniques based on the nutritional needs of plants, supplying their roots with a nutrient solution of minerals. In this paper, we implemented a DFT hydroponic system (Deep Flow Technique), which performs specific automated tasks to control the growth of plants, these tasks are: an algorithm capable of performing an automatic control, based on the methodology of fuzzy logic, where the system environmental parameters are regulated, using the antecedents, consequentials and inference rules given by the expert in the area. We developed a remote IoT-based monitoring of the state of plants germination using the segmentation of the seedling. The system requires the inter action of the IoT for monitoring the parameters obtained, using the MQTT  protocol to exchange information using a mosquitto broker-server on a Raspberry Pi 3 B+. In addition, we use an ESP32 microcontroller to exchange information with the broker-server to obtain the system parameters.  These parameters are observed in two interfaces, a mobile one and a server, to monitor the humidity in the environment, its potential hydrogen (pH), electrical conductivity, and the temperature of the nutrient solution.