Detection technology of plant protection equipment nozzle based on machine vision

摘要

In this paper, we present a method of using machine vision technology to detect the nozzle of plant protection equipment through two spray variables (spray angle and spray volume distribution area), which provides a good experimental direction to improve the performance of the nozzle of plant protection equipment. Firstly, we get the images information with different backgrounds and shooting angles. Secondly, we process the images by edge detection, mathematical morphology and other related digital image processing techniques. Thirdly, we obtain the target area of spray angle and fog quantity distribution using the Hough transform line detection algorithm and pixel method. Finally, we compare the area and angle with the reference values to judge the rationality of the method and establish the error range. Experimental results show that the feature variable values are ideal for plant protection equipment nozzle spray angle and spray volume distribution area. It is proved that the feasibility of our method.