Título:	Development of an image database of apple tree branches affected by european canker.
Autoria:	SILVA, E. C. da SESSI, A. S. MARCHIORETTO, L. de R. FACUNDO, B. R. PAULA FILHO, P. L. de GEBLER, L. ALVES, S. A. M.
Afiliação:	EDUARDO CARVALHO DA SILVA, EMBRAPA UVA E VINHO ALESSANDRA SOARES SESSI, EMBRAPA UVA E VINHO LUCAS DE ROSS MARCHIORETTO, EMBRAPA UVA E VINHO BRUNO RAPHAEL FACUNDO, UNIVERSIDADE TECNOLÓGICA FEDERAL DO PARANÁ PEDRO LUIZ DE PAULA FILHO, UNIVERSIDADE TECNOLÓGICA FEDERAL DO PARANÁ LUCIANO GEBLER, CNPUV SILVIO ANDRE MEIRELLES ALVES, CNPUV.
Ano de publicação:	2025
Referência:	In: WORKSHOP CIENTÍFICO DO CENTRO DE CIÊNCIA PARA O DESENVOLVIMENTO EM AGRICULTURA DIGITAL – SEMEAR DIGITAL, 2., 2025, Campinas. Anais [...]. Piracicaba: ESALQ/USP, 2025. p. 108-114.
Páginas:	7 p.
Conteúdo:	European canker is a significant disease affecting apple trees in Brazil. Manual detection is labor-intensive and time-consuming, with limitations in the early identification of lesions. This study aims to develop faster and more efficient detection methods through an automated system based on sensors or image processing. To achieve this, the creation of a comprehensive image database of affected apple tree branches is essential. The research was divided into two stages, both focused on building this database. The first stage involved compiling an RGB image dataset of healthy and infected branches. The second stage documented canker at various stages of development through a controlled inoculation experiment, using RGB and multispectral (725 nm) cameras. Preliminary results indicate that the physiological changes caused by infection produce detectable differences in the images, particularly at wavelengths above 700 nm. This underscores the potential of this spectral range for detecting diseased branches. The resulting database will enable detailed spectral analysis and will later be used to train convolutional neural networks (CNNs) to identify early infection patterns.
NAL Thesaurus:	Neonectria ditissima Spectroscopy
Palavras-chave:	Disease monitoring
Tipo do material:	Artigo em anais e proceedings
Acesso:	openAccess
Aparece nas coleções:	Artigo em anais de congresso (CNPUV)