The Architecture of Video Monitoring Systems in forest reserve

The issue of recognizing objects in the modern world does not lose its relevance. With the emergence and development of Deep Learning methods, the accuracy of object recognition has increased sufficiently. This in turn contributed to the rapid development of modern technology. The rapid development of technology has also facilitated the ability to perform complex computations on a low-cost device. Master's thesis is the aim of the architecture development of modern technologies (neural networks, deep learning algorithms, physical devices - Raspberry PI, WSN) which will reserves existing situation (animal variety of accounting, their population tvalqurisdevna and fire or other type of threat Rowley, to discover and to inform the relevant bodies). The master thesis provides a general overview of architecture. The main focus is on the development of an image and video data processing system using Python programming language. In particular, developing a CNN neural network on a Raspberry PI device and developing an appropriate system using machine learning techniques.