Object detection and tracking algorithms using brain-inspired model and deep neural networks

As the most effective bio-intelligence system, Human Visual System (HVS) has significant advantages in image processing, which helps to solve the problems in infrared target detection and tracking, such as dim small target, complex background, target occlusion and appearance changes, etc. In this paper, several brain-inspired models (including lateral inhibition, receptive field, synchronous burst, visual attention, and cognitive memory) and Deep Neural Networks (DNNs) have been studied, and the corresponding algorithms are proposed, which include: an infrared target detection algorithm based on lateral inhibition and singular value decomposition, an infrared target detection algorithm based on receptive field and lateral inhibition, an infrared moving dim target detection algorithm based on ALI-PCNN, an infrared target detection algorithm based on GCF-SB visual attention model, a kernel correlation filtering target tracking algorithm based on multi-channel memory model, and a robust and efficient discriminative-correlation-filter-based tracking approach based on the Response Map Analysis Network. Our experimental results show that the proposed algorithms are beneficial to achieve accurate infrared target detection and robust tracking under complex conditions.