Adaptive filtering method and device, electronic equipment, readable storage medium and chip

By using an adaptive filtering method to dynamically adjust the detection parameters of the millimeter-wave radar, the problem of insufficient detection accuracy in traditional methods is solved, and high-precision target identification is achieved in inland waterway environments.

CN120876911BActive Publication Date: 2026-06-09JIANGTONG (SHANGHAI) TECH CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Patents(China)
Current Assignee / Owner
JIANGTONG (SHANGHAI) TECH CO LTD
Filing Date
2025-07-17
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

When traditional millimeter-wave radar uses the constant false alarm rate (CFAR) detection method, it cannot achieve adaptive detection based on distance, time, and scene, resulting in insufficient target detection accuracy and problems of missed detections and false alarms.

Method used

By using an adaptive filtering method, the detection parameters are dynamically adjusted. Combined with the distance Doppler matrix data of the current frame and historical frames, the noise floor threshold and CFAR parameters are calculated in real time to perform two-dimensional detection, identify and suppress false target features, and improve the point cloud quality.

Benefits of technology

It effectively reduces false alarms and missed detections, improves the detection accuracy and reliability of millimeter-wave radar in inland waterway environments, and adapts to target identification in different navigation scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120876911B_ABST
    Figure CN120876911B_ABST
Patent Text Reader

Abstract

Embodiments of the present invention provide an adaptive filtering method, apparatus, electronic device, readable storage medium, and chip. The adaptive filtering method includes: acquiring range-Doppler matrix data of a millimeter-wave radar; determining a noise floor threshold for at least one range segment based on data from multiple consecutive frames; determining a point cloud effectiveness rate; determining an adaptive constant false alarm rate (CFAR) parameter for the current frame based on the point cloud effectiveness rate corresponding to historical frames; performing two-dimensional detection based on the adaptive CFAR parameter and the noise floor threshold for the current frame to determine an original point cloud; filtering invalid point clouds from the original point clouds corresponding to multiple consecutive frames to determine an optimized point cloud; identifying the scene corresponding to the optimized point cloud and determining false target features in the scene; and filtering the false target features in the optimized point cloud to determine a filtered point cloud. The solution of the present invention improves the accuracy of target detection by millimeter-wave radar.
Need to check novelty before this filing date? Find Prior Art