Reduction of Video Data at Translation of a Registered Object Relative to Video Sensors Based on the Eigenbasis of Interpretation Model
Dmitriy A. Balakin, Yu. P. Pyt’ev · Pattern Recognition and Image Analysis · 2022
Abstract The applications of the eigenbasis of a measurement interpretation model in the problem of reducing the measured video data to the form typical of measurements of the studied object by an ideal measuring device when the object is translated in the field of view of the video sensors are studied in this work. The methods for selecting the components in the eigenbasis are considered that allow to obtain an estimate of the studied object that has an error lower than that of the linear reduction of measurements by means of nonrandom distortion; the distortion is defined by the subspace of the values of the studied-object characteristic interesting to the researcher that does not influence the result. In addition, the variants for refining the position of the object for subsequent measurement reduction that correspond to different information about the object and the mathematical model of its motion are considered and studied.