Detection and tracking of weakly emitting point objects based on the analysis of a sequence of miniseries of images
Pavel A. Medennikov, Nikolai I. Pavlov · Journal of Optical Technology · 2021
This paper discusses the problem of detecting and tracking moving, weakly emitting point objects. It is proposed to solve it by using a sequential two-stage method of processing miniseries of images (frames) obtained by an optoelectronic system. The proposed method includes an original procedure of interframe processing of miniseries of images, a feature of which is to operate with two lists of objects: the main list and an intermediate (buffer) list. The indicated lists are updated after each miniseries of images is processed. The buffer list of objects in this case plays a subsidiary role and is used to filter out (eliminate) the false signals that appear during frame processing of the images. The use of the proposed interframe processing procedure makes it possible to work with low-SNR images, and this is regarded as a substantial advantage of the well-known track-before-detect method. Statistical experiments using miniseries of model frames confirm that the proposed method is workable.