Apparatus for identification and tracking of objects

Claim 1

Device for the acquisition and tracking of objects by means of characteristic features by which these objects are identifiable from a distance in the air in front of a background by means of an imaging sensor and a processing circuit for the signals obtained from the latter for comparison with reference signals stored beforehand, characterized by the combination of the following features: a) output signals of an imaging sensor (1), which is stabilized with respect to attitude, are analyzed with respect to brightness gradients contained therein for extracting edge lines (2) of the object; b) edge points are selected in a point selector (3) from the local extreme values of the brightness gradients by means of adaptive threshold value formation in the signal processing circuit; c) edge points lying on a straight line are represented by local extreme values by means of a parameter transformer (4), which local extreme values characterize a straight line portion by means of their parameters; d) the measured parameters are compared with the stored reference parameters (11) in an evaluating computer (5) while taking into account measured path data, and the results of the comparison are made available at its output for further processing; e) the output signal of the sensor (1) is analyzed within a plurality of image sections for average brightness and variance; f) a gray value histogram of edge points is calculated within a plurality of image sections; g) two threshold values which divide up the brightness values into three words, bright/reference/dark, and thus convert the image into a three-stage image are calculated from the data from e) and f) for each of the image sections; h) all image portions designated as "bright" or "dark" are roughly analyzed with respect to their magnitude, shape and position in order to determine whether or not they could be the sought target or whether or not possible target objects coincide with other objects and, if this is the case, the target objects are separated; and i) all possible target objects are accurately analyzed by means of shape, position and orientation features in order to determine whether or not they are sought targets.