During the transition of agriculture to the next stage, agriculture 5.0, artificial intelligence begins to play the central role. The key component is computer vision technology, which, combined with deep learning, enables agricultural productivity management and automation systems improvements to increase efficiency and profitability. Automation in agriculture provides benefits such as lower costs, high productivity, and precision. The paper considers the issues of using modern information technologies at the stage of agricultural machinery design, in particular, seeders for direct sowing of grain crops using a scatter method. Based on the electronic model created by the authors, a physical prototype was made using graphic editors. At the development stage, the authors justified the parameters of the seeding apparatus of a seeder for direct sowing of grain crops using a scatter method: a dispenser that provides the optimal number of seeds along the length of the sowing strip per 1 m2 of the sowing area; as well as a seed spreader, which ensures uniform distribution of seeds over a 3 m wide seeding strip and 660 mm disc sizes installed at angles: angle of attack in the range of 6° up to 12°, angle of lateral roll in the same range. These values were determined as a result of the experiment using an orthogonal central compositional plan of a factor experiment of a physical prototype of a seeder. The authors used technology for processing the results of individual experiments using computer vision, implemented by a program developed in Python using the OpenCV library of functions and computer vision tools, as well as the NumPy, Pandas, and Matplotlib libraries. The results of the final tests of the physical prototype of the seeder showed an even distribution of seeds, both along the length of the sowing strip and along its width. This ensured high seed germination, favorable conditions for plant development and a good yield of grain crops.