Computer vision based on machine learning technology: monograph

The monograph is devoted to the development of computer vision methods and tools using AI for fast identification of moving objects in a video data stream based on deep learning technologies. Classical and non-classical methods of artificial intelligence, convolutional neural networks, computer vision and image recognition, and theories of control systems based on estimates and criteria of mathematical statistics are considered. As a result of recognition, the type of recognised object is determined, and quantitative accuracy estimates are calculated. A method of applying templates is implemented. The algorithm has information about what the sought object looks like, what background it may have, what specific contours of the object look like and in what position they are. The possible location of the object detection is immediately taken into account. It allows for the achievement of high recognition quality and good performance. When a video camera shoots several similar objects, different templates are satisfied, and recognition decreases. Artificial neural network models are used to estimate or approximate functions that may depend on many inputs and are usually unknown.