Radar Detection Long Integration Kinematic Filtering

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current radar systems face challenges in detecting small maritime targets due to sea clutter, where high detection thresholds miss small targets and low thresholds result in false alarms, and existing kinematic filtering methods are computationally inefficient and specific to slow targets.

Innovation Solution

A method using long integration with kinematic filtering that groups detections from successive scans based on a uniform rectilinear movement model, employing a histogram to calculate temporal variations and test kinematic consistency, allowing for efficient detection of small targets regardless of speed and radar type.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the detection threshold is set to a high value to avoid false alarms from sea clutter, then the false alarm rate is reduced, but small targets with low RCS are not detected

Engineering Contradiction:
Improvefalse alarm rateVSAvoiddetection capability of small targets
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by performing long integration across multiple successive radar scans before making a detection decision. Detections from N successive scans are collected and integrated, allowing small targets to accumulate sufficient signal energy above the threshold while sea clutter remains random and does not consistently exceed the threshold across all scans. This preliminary integration of multiple scans resolves the contradiction by enabling small target detection without increasing false alarms.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If the detection threshold is set to a low value to detect small targets, then the detection probability of low RCS targets is increased, but the number of false alarms from sea clutter increases significantly

Engineering Contradiction:
Improvedetection capability of small targetsVSAvoidfalse alarm rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent uses preliminary integration of detections from N successive scans at a low threshold, then applies kinematic filtering as a second stage. The low threshold enables small targets to be detected in individual scans, and the long integration process accumulates these detections. The kinematic model then filters out false alarms by verifying consistency with expected target motion, resolving the contradiction between low threshold detection and false alarm rate.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces kinematic filtering as an intermediary step between low-threshold detection and final target confirmation. The kinematic model acts as a mediator that verifies whether detected targets follow physically plausible motion patterns, eliminating false alarms from sea clutter while preserving genuine small target detections. This intermediary process resolves the contradiction by adding a verification layer without requiring a high detection threshold.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Device complexity

If kinematic filtering uses a single coordinate (distance) to verify target consistency, then the process is simple to implement, but kinematic consistency is not strictly verified and computational efficiency is lost

Engineering Contradiction:
Improvefiltering process complexityVSAvoidkinematic consistency verification accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent extends kinematic filtering from single-coordinate (distance) verification to multi-coordinate verification by incorporating azimuth and elevation coordinates. The system verifies that detected targets are consistent across multiple spatial dimensions and time, checking whether position, velocity, and acceleration vectors satisfy kinematic models in 3D space. This dimensional extension strictly verifies kinematic consistency while maintaining computational efficiency through the histogram-based approach.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

4Measurement precision

If the kinematic filtering process uses multiple steps with specific tests and metrics, then detection accuracy is improved, but computational complexity becomes quadratic and the process is difficult to adapt

Engineering Contradiction:
Improvetarget detection accuracyVSAvoidcomputational efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces the traditional mechanical multi-step filtering process with a histogram-based computational approach. Instead of sequentially applying multiple tests and metrics that result in quadratic complexity, the system uses histograms to efficiently aggregate detection data across multiple scans and coordinates. The histogram bins store detection counts for different kinematic parameters, allowing rapid verification of kinematic consistency through simple counting operations rather than complex iterative tests, thus maintaining detection accuracy while achieving linear computational complexity.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3657200B1Detection method by long integration and corresponding radar system
Publication Date: 2023.08.30 THALES SA
  • EP3657200B1 patent drawingFigure 1
  • EP3657200B1 patent drawingFigure 2
  • EP3657200B1 patent drawingFigure 3

AI summary

At each new detection, called pivot detection, by a radar system, the process includes the steps of: grouping, with the pivot detection, grouped detections, defined as detections that belong to a scan preceding the sweep of the pivot detection and that have a non-zero probability according to a grouping criterion; filtering the grouped detections to retain only detections that are strictly kinematically consistent with the pivot detection, by: initializing (110) a histogram (H), each dimension of which is a time variation of a coordinate measured by the radar system; calculating (120, 130) an interval of potential values ​​for each coordinate of the pivot detection and each grouped detection; calculating (140) a minimum time variation and a maximum time variation for the coordinate(s) from the intervals of potential values ​​of the pivot detection and each grouped detection;incrementing (150) the set of histogram classes whose index according to each dimension lies between the calculated minimum and maximum temporal variations; and, detecting (160) a target as soon as at least one class of the histogram reaches a predefined value.;