DETECTION OF OBJECTS ON SPATIALLY INHOMOGENEOUS BACKGROUNDS USING NEURAL NETWORKS
A. K. Shakenov
Institute of Automation and Electrometry, Siberian Branch, Russian Academy of Sciences, Novosibirsk, Russia
Keywords: обнаружение и распознавание объектов, свёрточные нейронные сети, машинное обучение, малоразмерные объекты, object detection and recognition, convolutional neural networks, machine learning, small objects
Abstract
Several approaches to the use of neural networks for detecting objects on spatially inhomogeneous backgrounds are considered. A method for constructing a classifier for detecting objects directly from observed fragments has been implemented. An approach consisting of a combination of the matched linear filtering method and convolutional neural networks is proposed. It is shown that this approach reduces the likelihood of false alarms while maintaining the object detection probability.
|
