Development an artificial vision algorithm to detect the Alternaria Alternata disease in the citrus limon plant of the “Fundo Amada” (#816)

Authors

León León, Ryan Abraham

García Morales, Nathaly Nicolle

Tirado Palacios, Elia Teresa

Abstract

In the period from August to October 2023, lemon production in Peru was affected by unusual weather conditions linked to the El Niño phenomenon, generating a crisis. Factors such as the proliferation of pests and diseases, including penicillium, exocortis and mealy bug, as well as the threat of Alternaria Alternata, led to a decrease in availability and a 500% increase in the price of lemons. Faced with this scenario, Fundo Amada also experienced economic losses. To address Alternaria Alternata disease, an innovative approach was proposed by developing an artificial vision algorithm based on convolutional neural networks and Python. This algorithm demonstrated an efficiency of 95.8%, with only 5 errors out of a total of 120 samples, surpassing previous research with an accuracy of 98.3% and an effectiveness of 93.5%. The implementation of this system not only simplifies disease detection for farmers, but also lays the foundation for future research in agriculture and biological pest control. The visit to Fundo Amada validated the need for the project and highlighted its significant contribution to the development of innovative solutions to improve disease management in lemon plants and provide efficient responses to agricultural crises. This project stands out for its positive impact on the agricultural sector and its potential to drive future research in the field.