Valoración de cultivos de quinua, mediante teledetección y procesamiento de imágenes espectrales, para productores y comercializadores orgánicos bio taita Chimborazo (COPROBICH).

Abstract

The current research work was based on the investigation of artificial vision techniques and their use to assess quinoa crops' status through remote sensing and spectral image processing at Bio Taita Chimborazo Organic Producers and Traders Corporation (COPROBICH). The project involved obtaining quinoa crop images using a Micasense RedEdge Mx multispectral camera with remote sensing incorporated into the Phantom 3 drone to determine the levels of chlorophyll existing in the plant and from the percentages of absorption and reflection of the spectral bands determine the state of the plantation: healthy, usually healthy, sick or stressed, and dead, stones or bare soil. The process was developed in two stages: the first is the image acquisition stage that was carried out using the Phantom 3 drone, capturing the images from heights greater than 60 meters and in 5 different spectral bands: blue, green, red, infrared, and red border. The second stage is the treatment of the images that were developed in the SPIMCULT application, it is a system for processing multispectral images of cultures, and by calibration, segmentation, and filtration process, scaled images were obtained as a result of colors that represent the state of the crop based on a color bar. In the phenological stages of branching and panning, the application was tested in different quinoa crops in Riobamba canton, Chimborazo province. It is concluded that the results obtained were similar depending on the height of capture, as well as the application allowed a proper identification of the areas affected by the growing presence of pests and plant stress for better quinoa crop management.