Radar Target Tracker Moving Average Filter False Positives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Millimeter-wave radar systems face challenges in accurately detecting human motion while filtering out false positives from periodically moving objects, such as fans and cleaning robots, which can lead to unreliable target detection.
Innovation Solution
The implementation of a moving average filter in conjunction with moving target indicator (MTI) filtering to suppress range and Doppler bins with high moving averages, thereby generating an updated combined image that reduces false targets and enhances the detection of human/animal movements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional radar detection methods are used to detect all moving targets, then all motion can be detected, but false positives from periodically moving objects increase
Solution Approach 1:
The patent segments the detection process by dividing range-Doppler images into individual bins and processing each bin independently through the moving average filter. This segmentation allows selective suppression of only those bins exhibiting periodic motion patterns, while preserving bins containing genuine target motions. The detection reliability improves because false positives are filtered without affecting true target detection.
Solution Approach 2:
The patent applies preliminary action by computing moving averages over a sequence of historical range-Doppler images before performing suppression. This preliminary computation identifies bins with periodic motion patterns in advance, allowing the system to suppress only those specific bins in the current frame while maintaining sensitivity to genuine targets. The preliminary filtering action prevents false positives before they affect final detection.
2Object-generated harmful factors
If advanced filtering methods like vital sensing or machine learning are used, then false positive reduction improves, but memory and processing power requirements increase
Solution Approach 1:
The patent employs a computationally inexpensive moving average filter that requires minimal memory and processing resources compared to sophisticated methods like vital sensing or machine learning algorithms. The filter uses simple arithmetic operations on historical data, making it suitable for embedded platforms with limited resources. This disposable-like simplicity allows the system to achieve effective false positive reduction without the heavy computational burden of advanced methods.
Solution Approach 2:
The patent changes the detection parameter from complex pattern recognition (used in machine learning) to a simple statistical parameter - the moving average of detection values across multiple frames. By transforming the problem into computing and comparing moving averages, the system achieves false positive reduction with minimal computational complexity, suitable for resource-constrained embedded platforms.
3Measurement precision
If all range-Doppler bins are processed for motion detection, then comprehensive detection is achieved, but processing time and computational load increase
Solution Approach 1:
The patent extracts and processes only the necessary information by computing moving averages specifically for bins that show detection values across multiple frames. Instead of processing all bins uniformly, the system identifies and extracts only those bins requiring suppression based on their temporal patterns. This selective extraction maintains detection accuracy while reducing overall processing time and computational load.
Solution Approach 2:
The patent applies partial action by performing the computationally intensive moving average calculation and suppression operation only on bins that exhibit periodic motion patterns, rather than processing every bin in every frame. This partial processing approach maintains measurement precision for bins requiring suppression while improving productivity by skipping unnecessary processing of bins that do not require filtering.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides more reliable detection of human and animal motion by filtering out non-targets, using less memory and processing power compared to vital sensing and machine learning, and is suitable for embedded platforms without excessive power consumption.
Implementation Method 1
radar systems include a transmitting antenna for transmitting the radio-frequency (RF) signal, and a receiving antenna for receiving the reflected RF signal
Implementation Method 2
the distance between the radar and a target is determined by transmitting a frequency modulated signal, receiving a reflection of the frequency modulated signal (also referred to as the echo), and determining a distance based on a time delay and/or frequency difference between the transmission and reception
Data Source
AI summary
In accordance with an embodiment, a method includes: for each frame of a plurality of frames over time, obtaining a combined image by combining a plurality of range-Doppler images; for each range and Doppler bin in the combined images, accumulating values each associated to a given frame and indicative of whether a detection has occurred in the corresponding range and Doppler bin in the combined image of the given frame; for each range and Doppler bin in the combined images, determining a moving average based on the values accumulated over time for a corresponding range and Doppler bin; for a current frame, identifying range and Doppler bins in which repeated detection has occurred over past frames based on the moving averages, and generating an updated combined image for the current frame by suppressing the identified range and Doppler bins; and detecting motion in the updated combined image.