Eventos Anais de eventos
COBEM 2019
25th International Congress of Mechanical Engineering
ON THE INVESTIGATION OF ORDINARY SPRAY NOZZLE IMAGES TO DETECT ITS HEALTH CONDITION VIA IMAGE SIGNAL PROCESSING
Submission Author:
Gabriel Alves Costa , SP , Brazil
Co-Authors:
Gabriel Alves Costa, Marcos Silveira, Paulo José Paupitz Gonçalves, Michael Brennan, Fabricio Cesar Lobato de Almeida
Presenter: Gabriel Alves Costa
doi://10.26678/ABCM.COBEM2019.COB2019-1509
Abstract
Spray nozzles are commonly used in crops irrigation, with types of commercial nozzles available, selected according to the desired spray shape and application. This work intends to estimate the health condition of hollow cone nozzles, the type most used for spraying crops. Wear is the main mechanism which affects the spray pattern, jeopardizing effectiveness. Hence, image of sprays can be used to estimate the nozzle health condition, which is directly related to wear level. Ordinary side view images of spray nozzles are taken by using a GoPro camera, using a test bench developed by the company Spraying System Co. Moreover, the method herein developed to estimate the health condition of spray nozzles uses four pre-processed side view images taken from different positions. The pre-processing consists in turning the RGB image to greyscale, followed by getting a spray cross-section intensity vector near to the nozzle discharge, which is then filtered to attenuate the effects of undesirable noise and enhanced by applying the power-law transformation method. This gives how the intensity (amplitude) of the image changes along the pixels (spatial domain). The area under this plot is calculated for the four side view positions to estimate the health condition of the nozzle.
Keywords
nozzle wear, image processing, health condition, power-law transformation
