AUTOMAÇÃO DO PROCESSO DE CLASSIFICAÇÃO DA APTIDÃO AGRÍCOLA DO SOLO

Abstract

The classification of land agricultural suitability is a fundamental procedure for supporting land-use planning and soil management. It aims to identify the most suitable land use based on the physical, chemical, and biological attributes of the soil. In Brazil, the Land Agricultural Suitability Assessment System (SAA), proposed by Ramalho Filho (1995), is widely adopted but presents limitations due to the complexity and subjectivity involved in its application. To address these limitations, Pereira (2004) proposed a methodological revision that incorporates additional limiting factors and provides a more objective approach to classification. This study presents the development of an efficient classification algorithm, implemented using the Python programming language and based on Pereira's (2004) methodology. The algorithm was integrated into a mobile application developed with the Flutter framework, allowing its execution on multiple platforms. The tool enables the automatic processing of fieldcollected data, the identification of soil limitations, and the determination of agricultural suitability classes. The system aims to support agricultural professionals, such as farmers, technicians, and agronomists, by offering a precise, fast, and accessible classification process. Experimental evaluations demonstrate that the proposed algorithm is capable of producing accurate and consistent classifications, contributing to more informed decision-making in soil use and promoting sustainable land management practices.