Algorithms for contour detection in agricultural images

¹ Tashkent Institute of Irrigation and Agricultural Mechanization Engineers, National Research University, Tashkent, Uzbekistan
² Sejong University, South Korea, Seoul, Korea
³ Tashkent University of Information Technology after named Muhammad al-Khwarizmi, Tashkent, Uzbekistan

^* Corresponding author: m_narzullo@mail.ru

Abstract

Contour detection requires a more extensive use of them in agriculture. However, processing by experts on the basis of such images, including subjective visual observation of the field area, takes a lot of time and energy. One of the important parts of the image processing process is the problem of determining the contour of the object in the image, through which the objects in the image are extracted, that is, segmented. This research work is devoted to the comparative analysis of contour detection methods, and the presented methods were first tested on the basis of the original image and images whose contours were separated by experts. The test contour was performed based on the contour images extracted by the expert and the contour images generated by using the methods, and the comparison of the results was performed by the pixel comparison method. Based on the obtained results, an approach of applying the appropriate method depending on the quality of the image is proposed.