ANÁLISE TEXTURAL E CLASSIFICAÇÃO DIGITAL DE IMAGENS SAR PARA DISCRIMINAÇÃO DE ALVOS AGRÍCOLAS

Texture analysis and digital classification of SAR images for agricultural target discrimination

Autores

  • Silvio Pimentel MARTINS IBGE

DOI:

https://doi.org/10.5016/geociencias.v43i1.17919

Resumo

RESUMO - As imagens de sensoriamento remoto da faixa do visível e infravermelho do espectro eletromagnético apresentam grande potencial na discriminação de áreas agrícolas para fins de estimativa de safra. Contudo, a presença de nuvens impede a aquisição deste tipo de imagens, porém, as imagens SAR independem de condições meteorológicas. Neste contexto, este trabalho verificou o potencial de duas imagens SAR/Radarsat-1, banda C, polarização HH, nos modos Fine-5/ascendente (F5A) e Standard-7/descendente (S7D) na discriminação de alvos agrícolas na região de Assis-SP. Os métodos se basearam em análises visuais e na comparação das classificações digitais das imagens F5A e S7D originais, filtradas e medidas de textura destas. Os resultados indicaram que as imagens filtradas melhoraram a discriminação dos alvos em relação às imagens originais, sendo o filtro adaptativo Gamma, o mais eficiente entre os outros filtros testados. As classificações das imagens de textura foram, em geral, melhores do que as classificações das imagens filtradas, indicando que as medidas de textura podem ser atributos úteis para maximizar a discriminação de alvos agrícolas. As classes com maior potencial de discriminação em ambas as imagens F5A e S7D, com exatidão acima de 50%, foram: água, área urbana, cana-de-açúcar-2, soja e solo exposto-1. 

Palavras-chave: Radarsat-1. Máxima Verossimilhança (MAXVER). Filtros adaptativos. Sensoriamento remoto. Alvos agrícolas.

 

ABSTRACT - The remote sensing images of the visible and infrared range of the electromagnetic spectrum have great potential for discriminating agricultural areas for the purpose of estimating the yield. Nevertheless, the presence of clouds prevents the acquisition of this type of images, however, the SAR images are independent of meteorological conditions. In this context, this work verified the potential of two SAR/Radarsat-1 images, C band, HH polarization, in Fine-5/ascending (F5A) and Standard-7/descending (S7D) modes in the discrimination of agricultural targets in the region of Assis-SP. The methods were based on visual analyzes and on the comparison of the digital classifications of the original and filtered F5A and S7D images, as well as their texture measurements. The results indicated that the filtered images improved the discrimination of the targets in relation to the original images, with the Gamma adaptive filter being the most efficient among the other tested filters. Texture image classifications were generally better than filtered image classifications, indicating that texture measures can be useful attributes to maximize discrimination of agricultural targets. The classes with the greatest discrimination potential in both F5A and S7D images, with accuracy above 50%, were: water, urban area, sugarcane-2, soy and exposed soil-1.

Keywords: Radarsat-1. Maximum Likelihood (MAXVER). Adaptive filters. Remote sensing. Agricultural targets.

Biografia do Autor

Silvio Pimentel MARTINS, IBGE

Instituto Brasileiro de Geografia e Estatística (IBGE). Rua Coronel Peixoto, 17. Centro. Cuiabá – MT.

Downloads

Publicado

2024-04-06

Edição

Seção

Artigos