FIELD: radars.
SUBSTANCE: invention relates to methods of search and detection of an object on a terrain by a digital monochrome image of the terrain, for example, by radar image (RI) generated in synthetic antennae radars (SAR). Core of the method is that the entire search zone is divided into non-overlapping sections representing squares of search sized Ns×Ns pixels, in each of which one or more selective moments and/or coefficients of image brightness distribution are calculated and compared with threshold values. In case of exceeding one or several threshold values the decision is taken upon detection in this square of search a candidate for the searched object or its part and in all squares of search, where the decision on detection of the candidate is taken, its additional search is performed and its position is updated. As a selective moment and a selective coefficient of image brightness distribution the kurtosis is used. Size of the square of search Ns is selected comparable or less than length of the objects, but if provided a selection sufficient for calculation of selective coefficients of brightness distribution in the search window (Ns≥20). If in the SAR there are several channels with different resolution at detection of objects, at the stage of search RIs are used of a channel with lower resolution than the one at the stage of additional search. Additional search of the object is carried out within the range of squares of search, where the detection took place, in case of their expansion by the value of alignment error of RI of different channels and by the size of the square of detection. Additional search can be carried out by evaluating selective moments and/or coefficients of brightness distribution of RIs in the sliding window in the form of the square of detection sized Ndet×Ndet pixels. Herewith the size of the square of detection Ndet is selected smaller than the one of the square of search provided that the total number of independent selections of RI brightness values in the square of detection form a sufficient selection (Ndet≥20). Thresholds for each of the selective moments can be determined from an apriori statistics of these moments for fragments of RI background and RIs of detected objects with the background.
EFFECT: technical result is increased efficiency of detection of an object while reducing the amount of computations.
9 cl, 2 tbl
Title | Year | Author | Number |
---|---|---|---|
METHOD OF SEARCH AND DETECTION OF OBJECT | 2017 |
|
RU2648234C1 |
METHOD FOR LOCATION OF OBJECTIVE | 2006 |
|
RU2323450C1 |
METHOD FOR DETECTING AND DETERMINING COORDINATES OF SEARCH OBJECT | 2008 |
|
RU2392635C2 |
METHOD FOR SIMULATION OF GROUND OBJECT AT RADAR SURVEILLANCE | 2006 |
|
RU2310884C1 |
METHOD OF IDENTIFYING AND DETERMINING PARAMETERS OF IMAGE OF OBJECT ON RADAR IMAGES | 2012 |
|
RU2516000C2 |
METHOD FOR NAVIGATION OF AIRCRAFT BY RADAR IMAGES OF EARTH SURFACE WITH APPLICATION OF DIGITAL AREA MODELS | 2007 |
|
RU2364887C2 |
METHOD FOR INCREASING RADIOLOCATION RESOLUTION, SYSTEM FOR REALIZATION OF METHOD AND METHOD FOR REMOTE DETECTION OF SMALL OBJECTS BY SYSTEM | 2004 |
|
RU2265866C1 |
METHOD OF RECOGNITION OF SPACE APPLIANCES ON REDUCED RADAR LOCATION IMAGES | 2015 |
|
RU2620727C2 |
METHOD FOR AUTOMATICALLY RECOGNIZING SCENES AND OBJECTS IN AN IMAGE | 2021 |
|
RU2778906C1 |
METHOD FOR INTELLIGENT PROCESSING OF ARRAY OF NON-UNIFORM IMAGES | 2021 |
|
RU2767281C1 |
Authors
Dates
2016-09-10—Published
2015-06-16—Filed