The article deals with the creation of intelligent tractor driver support systems based on computer vision technologies for analyzing the direction of movement and detecting obstacles when performing specified operations, such as plowing, harrowing, weeding, and fertilizing. Electric power poles, trees, rocks, bird nests, animals, people, and field roads are identified as obstacles. The solution of functional problems in the system is based on the extraction of information from images using methods for detecting and recognizing objects in images. The analysis of existing approaches to solving the problems under consideration is carried out and it is shown that the use of deep neural networks is effective. The practical use of the methods based on the chosen approach is based on the performance of the computing system, the availability of sufficient training data and the optimality of the training method. It is shown that these factors are important when implementing an intelligent tractor driver support system.

No abstract

The purpose of this work is to develop an intelligent system for recognizing traffic signals. To achieve this, DetectNet was applied, using an interface for learning, which was developed by NVIDIA. With their help, the disadvantages of this approach were identified, and therefore it was necessary to consider another option for solving this problem.

No abstract

No abstract