Queued Radar Frame Feature Extraction for Sparse Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radar systems struggle to accurately identify both large and small objects, such as cars and pedestrians, due to sparse radar detections in individual frames, which leads to misidentification and inefficiencies in advanced driver assistance and autonomous driving applications.
Innovation Solution
A technique involving time-ordered queuing of radar frames, combined with temporal-spatial sampling and grouping, uses a hierarchical temporal-spatial encoder to extract object features by leveraging radar points across multiple frames, employing modified farthest point sampling and grouping based on both spatial and temporal proximity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar systems process individual frames independently, then processing speed is maintained, but object identification accuracy deteriorates due to sparse radar detections
Solution Approach 1:
The patent transitions from processing single 2D radar frames to processing 3D temporal-spatial volumes by stacking multiple frames in a queue. This dimensional extension allows the system to leverage temporal information across frames, improving object detection accuracy for small objects like pedestrians while maintaining processing efficiency through volumetric processing architecture.
2Reliability
If radar systems use temporal-spatial processing across multiple frames, then object representation density improves, but computational complexity increases
Solution Approach 1:
The patent segments the temporal-spatial processing into distinct functional components: frame queuing module, volumetric encoding module, and feature extraction module. This segmentation allows each component to be optimized independently, managing computational complexity while achieving robust object representations through multi-frame temporal-spatial analysis.
Solution Approach 2:
The system performs preliminary actions by pre-processing and queuing radar frames before volumetric encoding. Frames are organized and prepared in advance, allowing the main processing pipeline to operate more efficiently on structured data, thereby reducing overall computational complexity while maintaining representation robustness.
3Reliability
If radar systems process only current frame data, then processing efficiency is maintained, but detection reliability for small objects deteriorates
Solution Approach 1:
The system performs preliminary action by maintaining a queue of historical radar frames and pre-computing volumetric representations. This allows the detection algorithm to access accumulated temporal information without real-time processing delays, improving small object detection reliability while minimizing processing time through advance preparation.
Data Source
AI summary
A computerized technique is disclosed of identifying object features in an environment of a vehicle. The technique includes receiving, by an encoder, data representing a plurality of frames, the frames providing point-in-time versions of a segmented pointed cloud derived from output of one or more radar sensors of the vehicle and including points that represent radar detections corresponding to an object in the environment at respective instants in time. The technique further includes arranging the plurality of frames in a time-ordered queue and processing the frames in the queue, including (i) selecting, from among the points, a plurality of sample points that spans multiple frames of the queue, (ii) forming a plurality of groups of points based on respective sample points of the plurality of sample points, and (iii) extracting features of the object based on the plurality of sample points and the plurality of groups.