The article describes innovative technologies for training engineering specialists. The main meaning of improving the future specialists’ professional training is that the student not only accumulates knowledge and acquires the necessary skills during this period, but also harmoniously develops all components of professional competence. However, at the same time, they are also two relatively independent processes that have both common and special principles in the future specialist personality development single process.

The article deals with the design and software implementation of neural network modules for solving problems of image recognition. In particular, it describes the development of a module for network training and recognition of input pulses, which made it possible to recognize, process and analyse aerial photographs of agricultural crops as objects of identification based on the use of a multi-layer deep learning neural network. The practical use of the software tool is possible for the study and study of the peculiarities of cultivation in conditions of irrigation and differentiated placement.

The article discusses the possibilities of using the mathematical and instrumental enhancement of the multilayer Hamming network in the tasks of recognizing aerial photographs, in particular, the advantages of the neuromodel for implementing further research to solve the problems of increasing crop yield based on monitoring and segmentation of dry crop areas are highlighted.