Approximation of an Image Formed for a Small-Size Moving Object by a Scanning Multirow Photodetector

The possibility to improve the precision of measuring the coordinates and brightness of small-size moving objects in a sequence of images formed by a scanning multirow photodetector is studied. Precision is increased by approximating the image of an object by a function, whose parameters are the brightness of the object and the shift of the center of its projection relatively the array of photosensitive cells (PSC) in the detector. The approximation quality criteria are, on the one hand, the mean-square difference between the image of the object and the approximating function and, on the other hand, the spread of estimates for its coordinates and brightness depending on the position relatively the center of a photodetector cell in a series of images. Several most common functions, which are usually used to approximate the discrete images of small-size objects, are compared in a series of images taken by projecting an illuminated small-diameter diaphragm in the plane of the photosensitive layer of the detector. It is shown that, even at a rather high level of random noise, the application of an approximating model makes it possible to estimate the coordinates of an object with a mean-square error less than 20 $$\%$$ from the PSC array mesh size.