NLOS Correction for Radar Target Detection in Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving systems face challenges in accurately detecting and classifying objects in the surrounding environment, particularly due to Non-Line-of-Sight (NLOS) reflections from planar surfaces, which can lead to incorrect target identification and localization, especially in all-weather conditions.
Innovation Solution
The implementation of a radar system with a meta-structure (MTS) antenna capable of steering beams in a 360° field of view, combined with a NLOS correction module that uses a Perception Module and reinforcement learning engine to generate a corrected point cloud, and sensor fusion with other sensors like cameras and lidar to enhance object detection and identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional radar systems are used for target detection, then the system structure is simple, but NLOS reflections cause incorrect target identification and localization
Solution Approach 1:
The patent introduces an NLOS correction module as an intermediary component between the radar sensor and the target detection system. This module receives the point cloud data, identifies NLOS reflections by analyzing geometric relationships with planar surfaces, and corrects the target locations by computing virtual source positions. The intermediary module resolves the contradiction by adding only the necessary complexity to eliminate NLOS errors without redesigning the entire radar system.
Solution Approach 2:
The patent segments the target detection process into distinct functional modules: the original radar detection system and the added NLOS correction module. The correction module further segments the problem by separately identifying planar surfaces, detecting NLOS reflections, and computing corrections. This segmentation allows the system to address NLOS issues independently without complicating the overall system architecture.
2Measurement precision
If NLOS correction processing is added to eliminate reflection errors, then target identification accuracy improves, but processing time and computational complexity increase
Solution Approach 1:
The patent applies preliminary action by first identifying and storing the geometric parameters of planar surfaces in the environment before performing target detection. When NLOS reflections are detected, the system uses pre-computed surface geometries to quickly determine correction positions rather than performing complex geometric calculations in real-time. This preliminary preparation significantly reduces the processing time required for NLOS correction.
3Reliability
If multiple sensors are integrated for sensor fusion, then detection accuracy in all-weather conditions improves, but device complexity and cost increase
Solution Approach 1:
The patent implements sensor fusion where a single radar system performs multiple functions: traditional target detection and NLOS reflection correction. The NLOS correction module uses the same radar point cloud data for both detecting actual targets and identifying reflection artifacts, eliminating the need for separate correction sensors. This multi-functionality approach maintains all-weather detection reliability while avoiding the complexity of integrating additional specialized sensors.
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 solution enables accurate and efficient detection and classification of objects in all-weather conditions, reducing processing time and computational complexity, and providing a human-like interpretation of the environment, enhancing the capabilities of autonomous driving systems.
Implementation Method 1
the ability to detect and classify objects in the surrounding environment... may include a combination of multiple sensors, such as camera, radar, and lidar
Implementation Method 2
Non-Line-of-Sight (NLOS) reflections from planar surfaces, which can lead to incorrect target identification and localization
Data Source
AI summary
Examples disclosed herein relate to an autonomous driving system in a vehicle having a radar system with a Non-Line-of-Sight (“NLOS”) correction module to correct for NLOS reflections prior to the radar system identifying targets in a path and a surrounding environment of the vehicle, and a sensor fusion module to receive information from the radar system on the identified targets and compare the information received from the radar system to information received from at least one sensor in the vehicle.