Determination of a computational parameter of quantitative production of crops in hydroponic greenhouses through digital image processing

摘要

The objective of this paper is to determine a computational quantitative production parameter of tomato crops through digital image processing and neural networks. A series of computational processes are performed in order to calculate the total pixel area of the tomatoes displayed, and to obtain a quantitative indicator of hydroponic crop production. It was possible to identify, at the pixel level, objectives belonging to hydroponic tomatoes with an error rate of 1,07%, and based on that, an indicator of quantitative production was obtained through computational vision.