Aplicativo mobile de detecção de praga de lagartas e Diabrotica speciosa em lavouras de soja

Abstract

This work addresses the development of a mobile application based on Artificial Intelligence (AI) for pest detection in soybean plantations. The justification for this project lies in the importance of agriculture to global economies and the need to combat pests that significantly impact agricultural production, such as caterpillars and the Diabrotica speciosa beetle. The objectives include a literature review on caterpillar and beetle pests, an investigation of emerging technologies applicable to pest detection (with an emphasis on AI), and the design and development of the mobile application. The study highlights the feasibility of AI, particularly Machine Learning, in pest detection through image analysis, as demonstrated in previous studies. The use of REST APIs and Google Cloud AutoML Vision are explored as key tools for the application's implementation. This research aims to democratize access to pest detection technologies in an accessible and effective manner, contributing to food security and the sustainable growth of agriculture. The results obtained from the application’s development were promising, demonstrating the effectiveness of artificial intelligence in detecting pests in crops. During testing, the system achieved a high accuracy rate in identifying caterpillars and the Diabrotica speciosa beetle when trained with 1,500 images from the image database for each pest type, despite the high implementation cost. This allows farmers to quickly and reliably access information about infestations in their crops.