Detection and validation of drops generated by agrochemicals application

Agrochemicals use has grown, from the 1980s onwards, proportionally to the development of agriculture, both in Argentina and worldwide. This growth is more noticeable in the case of herbicides, which present an undeniable benefit in food production, providing larger and more stable production. However, the application of these products in an improper and careless way carries possible toxic risks to health and the environment, if too high, or its effectiveness may be reduced, if too low. This need of controlled application raises the issue of the measurement of the distribution of the chemicals as they are being applied. One of the most used resources for this task is the hydrosensitive card (HC), which proved to be an effective tool for the analysis of the “quality of application” of agrochemicals in a field. The efficiency of these cards depends strongly on a precise readout of the shape, quantity and distribution of the droplets that hit the card, when placed on the application area. Much work is actually devoted to this task, but most of the tools are not available for field work, and the few ones that are available are not customizable. In this work we developed an algorithm for the lecture and analysis of hydrosensitive cards. It detect automatically the drops, determines the quality of the card, and measures several statistical quantitative parameters, useful for the control of the method and quantity of application.