MONITORING OF SOYBEAN RUST THROUGH IMAGES

Resumo

The automatic occurrence of quality in agribusiness is becoming essential in order to reduce costs and
improve quality. Using image analysis techniques, it is possible to carry out pest control in various crops, such as
soybean plantations, by obtaining aerial images, for example using drones. An image when scanned has,
depending on its format, that is, its construction method, three layers called RGB (Red - Green - Blue) that
precede the final image to be demonstrated on monitors, displays and printers. Concepts that the construction of
the image depends on the type of sensor used and considering a digital image sensor capable of capturing the
luminosity of the images that are projected on it continuously, we have the beginning of the process of capturing
a moment of time, called photographs, or for a sequence of images over time. For capturing images in cores, it is
common for video cameras to use three types of sensors, known as the 3CCD system, where each sensor has a
tri-conical filter on it, and photographic cameras generally have a single image sensor that groups your photo
sites under a mosaic of luminosity and color filters. The computational tool developed will allow the soy grower
to monitor and identify the appearance of rust in the early stages and will avoid the unnecessary use of
fungicides, reducing costs and reducing environmental impacts.