Modern approaches to image segmentation in agriculture

¹ Siberian Federal University, Krasnoyarsk, Russia
² Krasnoyarsk State Agrarian University, Krasnoyarsk, Russia
³ Reshetnev Siberian State University of Science and Technology, Krasnoyarsk, Russia
⁴ Navoi State University of Mining and Technologies, Navoi, Uzbekistan
⁵ National Research University “Tashkent Institute of Irrigation and Agricultural Mechanization Engineers”, Tashkent, Uzbekistan
⁶ FSBEE HE Siberian Fire and Rescue Academy EMERCOM of Russia, Zheleznogorsk, Russia

^* Corresponding author: greys.ameliya@mail.ru

Abstract

Image segmentation is one of the key areas in computer vision, as it allows for the identification and isolation of distinct regions, objects or structures within an image, which is critical for subsequent analysis and processing of visual data. This article discusses the fundamental principles, capabilities and limitations of various segmentation methods. Special emphasis is placed on the use of the Python programming language, which, thanks to its rich ecosystem of libraries such as OpenCV, TensorFlow, PyTorch, and scikit-image, has become the standard tool for the development and implementation of computer vision algorithms. The prospects for further development of segmentation technologies are discussed in the context of increasing data volumes and increasing requirements for the accuracy and efficiency of analysis. In the article, practical examples of applying segmentation models in agriculture are also presented.