Background adaptive target detection and tracking with multiple observation and processing stages
First Claim
1. A scanning electromagnetic sensor apparatus, comprising:
- a focal plane assembly comprising a plurality of detectors arranged substantially radially and each producing an electrical signal in response to electromagnetic radiation;
means for rotating an electromagnetic radiation emitting image with respect to the focal plane assembly;
means for subtracting a background signal from each electrical signal,wherein the subtracting means includes a clutter-adaptive algorithm that varies as a function of background characteristics;
means for filtering the electrical signal to produce a filtered signal from each detector after the background signal has been subtracted using an in-scan and cross-scan filtering algorithm to maximize signal-to-noise ratio;
means for generating an adaptive threshold value as a function of residual noise standard deviation;
means for determining whether each filtered signal exceeds the adaptive threshold value;
means for determining whether each filtered signal that exceeds the adaptive threshold value is a peak.
2 Assignments
0 Petitions
Accused Products
Abstract
A computationally efficient missile detection and tracking system, operating on data samples from a focal plane assembly (FPA) of an electro-optical sensor which is observing the earth and space, detects missiles, aircraft, resident space objects or other objects in the presence of background clutter, determines missile types and estimates missile position and velocity components in an inertial reference frame. Digitized samples from an array of electro-optical detectors which are viewing the earth and/or space are processed to detect and track missiles, aircraft and resident space objects. Streak detection is used for this purpose, where the target is separated from a background scene by exploiting target motion relative to the background scene. This processing system provides superior performance and simplified implementation as compared to conventional methods. The weight and power of a computer which can implement the processing system is much reduced as compared to conventional implementation.
-
Citations
44 Claims
-
1. A scanning electromagnetic sensor apparatus, comprising:
-
a focal plane assembly comprising a plurality of detectors arranged substantially radially and each producing an electrical signal in response to electromagnetic radiation; means for rotating an electromagnetic radiation emitting image with respect to the focal plane assembly; means for subtracting a background signal from each electrical signal, wherein the subtracting means includes a clutter-adaptive algorithm that varies as a function of background characteristics; means for filtering the electrical signal to produce a filtered signal from each detector after the background signal has been subtracted using an in-scan and cross-scan filtering algorithm to maximize signal-to-noise ratio; means for generating an adaptive threshold value as a function of residual noise standard deviation; means for determining whether each filtered signal exceeds the adaptive threshold value; means for determining whether each filtered signal that exceeds the adaptive threshold value is a peak.
-
-
2. A system for detecting and tracking objects when given data from a focal plane assembly of an electro-optical sensor, comprising:
-
an algorithm for rejecting spurious data spikes; an algorithm for correcting for detector responsivity error; an algorithm for correcting for detector offset error; an algorithm for performing time delay integration; an algorithm for correcting for focal plane geometry; an algorithm for correcting for non-uniform data sampling effects; an algorithm for performing clutter-adaptive estimation and subtraction of a scene background; an algorithm for performing matched filtering; an algorithm for performing target detection with a constant false alarm rate by setting a clutter-adaptive detection threshold; an algorithm for forming two-dimensional data packets of threshold-exceeding data for transmission to a programmable processor; an algorithm for specifying window parameters for scene windows; an algorithm for sending data from scene windows to an object dependent processor; an algorithm for rejecting clutter points using a pulse shape test; one or more algorithms for calibrating sensors; an algorithm for correcting digital logic approximations; an algorithm for correcting sensor jitter using observed ground sources; an algorithm for converting line-of-sight measurements in sensor coordinates to line-of-sight measurements in inertial reference frame coordinates using sensor attitude measurements; an algorithm for classifying observed objects to distinguish earth-fixed clutter points from missiles; an algorithm for identifying missiles and rejecting false missile tracks by calculating energy per unit mass of observed objects; an algorithm for adaptively varying a threshold multiplier used in the target detection algorithm to maintain a constant false alarm rate in severe clutter; an algorithm for defining the scene windows, centered on selected objects being tracked for identifying extended objects; an algorithm for line-of-sight tracking of missiles, wherein the observations of a given missile are smoothed to produce, at a specified epoch time, a unit line-of-sight position, velocity and acceleration vectors; an algorithm for using a priori knowledge of missile boosting profiles of observed objects to calculate missile position and velocity at burnout; an initial orbit determination algorithm for calculating missile position and velocity during a post-boost phase of flight; and an algorithm for using the line-of-sight position, velocity and acceleration vectors to calculate commands for pointing a remote electro-optical sensor toward selected objects. - View Dependent Claims (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A method of detecting and tracking objects using data received from passive optical sensors, including a scanning acquisition sensor capable of sensing signals emitted in an acquisition sensor field-of-regard, and a tracking sensor capable of sensing signals emitted in a tracking sensor field-of-regard, the method comprising the steps of:
-
separating likely target data received from the acquisition sensor from spurious data by utilizing one or more background-adaptive detection algorithms; reducing false detections resulting from the step of separating likely target data samples by using multiple scans of the acquisition sensor field-of-regard to identify earth-fixed objects and rejecting data samples which are not consistent with known object motion and intensity characteristics; initiating tracks of objects of interest by storing line-of-sight vector data associated with the objects of interest; reducing false detections by testing a hypothesis that the tracks of objects of interest are consistent with predetermined characteristics of selected object types; scheduling an observation time for the tracking sensor to observe specified objects being tracked; calculating a sequence of line-of-sight pointing commands to cause the tracking sensor to be slewed to a sequence of positions for observing each specified object during the scheduled observation time intervals; using a background-adaptive streak detection algorithm to detect the specified objects in the tracking sensor field-of-regard; combining the data received from the acquisition sensor data with the data received from the tracking sensor data to form tracks; and controlling background-adaptive detection algorithm parameters so as to adapt to operating conditions. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39)
-
-
40. A method of detecting and tracking objects using data received from a passive optical sensors, the sensor capable of sensing signals emitted in a sensor field-of-regard, the method comprising the steps of:
-
separating likely target data received from the acquisition sensor from spurious data by utilizing one or more background-adaptive detection algorithms; reducing false detections resulting from the step of separating likely target data by using multiple scans of the sensor field-of-regard to identify earth-fixed objects and rejecting data which are not consistent with known object motion and intensity characteristics; initiating tracks of objects of interest by storing line-of-sight vector data associated with the objects of interest; reducing false detections by testing a hypothesis that the tracks of objects of interest are consistent with predetermined characteristics of selected object types; scheduling an observation time for the sensor to observe specified objects being tracked; calculating a sequence of line-of-sight pointing commands to cause the sensor to be slewed to a sequence of positions for observing each specified object during the scheduled observation time intervals; using a background-adaptive streak detection algorithm to detect the specified objects in the sensor field-of-regard; controlling background-adaptive detection algorithm parameters so as to adapt to operating conditions.
-
-
41. A method for subtracting a background signal sensed by an electromagnetic sensor that includes a plurality of detectors, using a background subtraction algorithm, the method comprising the steps of:
estimating the background that each detector is observing by recursively filtering data output by each detector in order to subtract that estimate from each detector output, wherein an effective length of an averaging function implicit or explicit in the recursive background estimation filter is adapted to a local background by means of a steepest descent method which minimizes the variance of a data residuals output by the background subtraction algorithm. - View Dependent Claims (42, 43, 44)
Specification