Range profile target classifier
First Claim
1. A range profile target classifier for identifying targets according to a sequence of received range profiles observed at a plurality of aspect angles, said aspect angles having a predetermined aspect angle increment, comprising:
- input means for receiving a multiplicity of discrete range profile data signals, each of said multiplicity corresponding to a discrete look at said target;
a plurality of profile correlator means coupled to said input means for comparing said received discrete range profile data signals to a plurality of reference profile data signals, each of said profile correlator means corresponding to a predetermined target and each of said reference profile data signals corresponding to a preselected aspect of said predetermined target and for providing a correlation number representative of match between said received and said reference profile data signals;
a plurality of sequence correlator means coupled to a corresponding one of said plurality of profile correlator means for providing a signal representative of the maximum sum of said correlation numbers obtained from successive looks at said target, said sums selected from looks having aspect angles changing by no more than an integral number of said predetermined aspect angle increment and having no more than one of said aspect angles for each of said looks; and
maximum detector means coupled to said plurality of sequence correlator means for detecting the largest of said maximum sum signals received from said plurality of sequence correlator means and for providing a signal representative thereof.
1 Assignment
0 Petitions
Accused Products
Abstract
A range profile classifier wherein targets are identified based on the sequences of their received range profiles. Individual looks at a target are taken and it is assumed that the aspect angle of observation changes by no more than a predetermined amount. Correlations are made with reference data, which correlations are then summed recursively according to preselected restrictive criteria to obtain the most likely sequence corresponding to the target for which reference data is stored in the range profile target classifier.
55 Citations
13 Claims
-
1. A range profile target classifier for identifying targets according to a sequence of received range profiles observed at a plurality of aspect angles, said aspect angles having a predetermined aspect angle increment, comprising:
-
input means for receiving a multiplicity of discrete range profile data signals, each of said multiplicity corresponding to a discrete look at said target; a plurality of profile correlator means coupled to said input means for comparing said received discrete range profile data signals to a plurality of reference profile data signals, each of said profile correlator means corresponding to a predetermined target and each of said reference profile data signals corresponding to a preselected aspect of said predetermined target and for providing a correlation number representative of match between said received and said reference profile data signals; a plurality of sequence correlator means coupled to a corresponding one of said plurality of profile correlator means for providing a signal representative of the maximum sum of said correlation numbers obtained from successive looks at said target, said sums selected from looks having aspect angles changing by no more than an integral number of said predetermined aspect angle increment and having no more than one of said aspect angles for each of said looks; and maximum detector means coupled to said plurality of sequence correlator means for detecting the largest of said maximum sum signals received from said plurality of sequence correlator means and for providing a signal representative thereof. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method for identifying targets according to a sequence of received range profiles observed at a plurality of aspect angles, said aspect angles having a predetermined aspect angle increment, said method comprising:
-
receiving a multiplicity of discrete range profile data signals, each of said multiplicity corresponding to discrete look at said target; comparing said received discrete range profile data signals to a plurality of reference profile data signals, each of said reference profile data signals corresponding to a preselected aspect of said target and further providing a correlation number representative of match between said received and reference range profile data signals; providing a signal representative of the maximum sum of said correlation numbers obtained from successive looks at said target and corresponding to aspects changing by no more than an integral number of said predetermined aspect angle increment, and having no more than one of said aspect angle for each of said looks; and detecting the largest of said maximum sum signals and providing a signal representative thereof. - View Dependent Claims (11, 12, 13)
-
Specification