Smartphone image-based methods for quick, non-invasive assessments in agriculture

• Autor(a) principal: Carreira, Vinicius dos Santos [UNESP]
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://hdl.handle.net/11449/237462
• topic: Processamento de imagens - Técnicas digitais; Agricultura de precisão; Visão por computador

Image processing timely contributes for agricultural purposes. Its outstanding benefit of extracting information from the field can replace timeconsuming, labor-intensive methods. However, most times innovative methods use high-complexity systems and expensive equipment. Low-income regions are not able to finance such costs. Therefore, we concentrate efforts to develop smartphone imagebased methods for agricultural measurements using a mobile device. Smartphone are widespread technology known as flexible, affordable systems. To prove our hypothesis that low-complexity approaches can be useful, we analyzed two agricultural operations. In Chapter 2, we describe how two image-based methods (Region of Interest and Local Maxima) can detect maize plants and estimate their uniformity with overall results of RMSE of 3.98 cm and R² of 0.90. In Chapter 3, we created a smartphone image-based framework to measure macro spray characteristics, including spray angle, distribution shape and solid stream jets. Using image enhancement and morphological operations, our framework accurately resembles traditional methods and can be useful for quick screening of spray nozzles. Therefore, we prove that smartphone use can go beyond ordinary activities and help to spread digital assessments in agriculture.