Benefits of using spraying software

Comparison of three different software used to analyze images of spray droplets shows differences and similarities in the results obtained.

In phytosanitary control, the effectiveness of pesticide application is closely related to the quantity and way in which the applied product reaches the target. Based on this, the use of software for spectrum and droplet population analysis is a fundamental tool for quantifying and classifying the volumetric deposition of the applied spray. In general, these programs help in making decisions about the quality of the application, and there are several paid and free software options on the market aimed at this task.

Among them, we can mention CIR 1.5, which is a paid program and already renowned in the market, and Image Tool, which is a free tool, both developed for the Windows operating system. Furthermore, the Gotas application has been used in a practical way, which is free and developed for the Android mobile operating system. This application stands out mainly for its mobility and practicality, and can be used on smartphones and tablets. Despite the advantages offered by the different options available to producers and technicians, it is noted that, in practice, programs may present differences in their results. This can cause distortions in decision-making regarding the quality of pesticide application, which demonstrates the importance of analyzing the software used.

To try to find possible differences between the results of each program and size them, three software used to analyze images of spray droplets were evaluated: CIR 1.5 (Conteo y Tipificación de Impactos de Pulverización), Image Tool 3,0 and the Drops app 1.2. 16 water-sensitive labels (26mm x 76mm) sprayed with water were used, using the following tips: API-11002 (single fan jet-JLS), AD/D-11002 (double fan jet-JLD), AVI TWIN-11002 ( double fan jet with air induction-JLDIA) and APM-09502 (single impact fan jet-JLSIP), with four repetitions.

Spraying was carried out in the absence of wind, using a hydraulic CO backpack sprayer.₂ pressurized, operating at a working pressure of 300kPa, applying 200L/ha, and with a spray bar height of 0,5m. The spots on the labels were analyzed in each of the three software programs and the following parameters of the droplet spectrum produced were determined: the volumetric median diameter (DMV), the dispersion coefficient (SPAN) and the coverage density (DEN). The volumetric median diameter (VMD) is the diameter of the droplet that separates the total set of drops produced by spraying the syrup into two halves with the same amount of liquid. Therefore, the sum of the volume of drops with a diameter smaller than the DMV is equal to the sum of the volume of drops larger than the same. The dispersion coefficient (SPAN) measures the relative amplitude that exists between the diameter of all the drops produced, and the closer its value is to zero, the greater the degree of uniformity of the drops. And the coverage density (DEN) is measured through the number of drops deposited in each cm² of the pulverized target.

Immediately after application, the labels were collected and the volumetric median diameter, dispersion coefficient and coverage density were determined in the Drops application. Then, the labels were digitized (1.200dpi) and these same parameters were evaluated in the CIR and Image Tool software. The data obtained were subjected to analysis of variance (Anova) and a subsequent mean test (Tukey-5%).

The Gotas application, despite stopping working in some situations during the analysis process, proved to be quick and practical in evaluating the stain image. The first step to analyzing the label in this software is capturing the image, this must be done as quickly as possible to prevent the label surface from darkening due to the effect of humidity in the air. The label must be placed on a flat surface and must completely fill the area indicated in the application, so that the system can calibrate the droplet size, taking the standardized dimensions of the label as a reference. The second step is to select the edit option and go to crop - you can also use the highlight option before cropping, to improve the image quality. The third step is to select an area of ​​the label that best represents the application. This area should be as large as possible, but the edges of the label or areas where there are smudges should be avoided. In the fourth and final step, the application will show the area that was previously selected. You must now click on the process option and then the results will be shown.

The CIR and Image Tool programs are less practical, as they require the labels to be scanned in advance. However, from this point onwards, CIR stands out in relation to Image Tool, as it performs the calculations automatically, giving the results in tabulated form. Image Tool requires the user to perform analyzes manually and calculate the final data in a separate table.

COMPARATIVE

The analysis of variance of the data was significant for the three characteristics (volumetric median diameter, dispersion coefficient and coverage density). In other words, the programs presented different values ​​for all of them. However, there was no difference between the CIR and Gotas software for the volumetric median diameter. On the other hand, the dispersion coefficient was different in the three programs. In contrast, the coverage density was similar in the Drops and Image Tool applications. The Image Tool program tended to have much higher values, on average up to 84% higher for the volumetric median diameter (Table 1). Possibly because it is difficult to distinguish very small spots on the surface of the labels. Some studies already carried out have shown that this program, in relation to other tools, is not efficient in determining these values. However, for the dispersion coefficient and coverage density, this same program presented median values ​​(Table 1).

These same studies observed that the Image Tool is an excellent tool for analyzing coverage density. In relation to the Gotas application, this resulted in median values ​​for the volumetric median diameter and values ​​on average up to 28% and 81% lower for the dispersion coefficient and coverage density, respectively (Table 1). On the other hand, the CIR software presented values ​​on average up to 46% lower for volumetric median diameter and 39% and 419% higher for dispersion coefficient and coverage density, respectively (Table 1).

The ability of this program to identify a greater number of drops/cm^-2 It is a characteristic already observed in other situations. This is due to the greater sensitivity of this software in distinguishing the overlap between drops. Furthermore, good consistency of the data from this application has already been observed when evaluating intact water-sensitive labels.

Finally, it can be concluded that there was an overestimation in the determination of droplet size by the user when using the Image Tool 3.0 software, compared to CIR 1.5 and Gotas 1.2.

João de Deus Godinho Júnior, Guilherme Andrade Gontijo, Renato Adriane Alves Ruas, José Márcio de Sousa Júnior, Waner Gleider Barbosa, UFV

Article published in issue 169 of Cultivar Máquinas