Radar Data Recognition Using Spatially Variant Input Configurations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radar data recognition methods face challenges in accurately classifying objects and verifying their authenticity using spatially invariant input data, which limits their ability to distinguish between real and fake objects effectively.
Innovation Solution
A processor-implemented method that generates multiple pieces of input data with different dimension configurations based on radar data, using a recognition model to extract feature data through convolution filtering and fuse it for generating a recognition result, including authenticity information, by considering the correlations between angle, range-velocity, and time frame components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If spatially invariant input data is used for radar data recognition, then the processing is simpler, but the accuracy of object classification and authenticity verification deteriorates
Solution Approach 1:
The patent transforms the radar data from spatially invariant representation to spatially variant representation by introducing multiple dimension configurations (angle dimension, range-velocity dimension, time frame dimension). This dimensional transformation enables the system to capture spatial relationships and contextual information that were lost in the original spatially invariant data, thereby improving recognition accuracy without excessive complexity increase.
Solution Approach 2:
The patent segments the radar data into multiple pieces of input data, each with different dimension configurations. By dividing the data into separate components (angle component, velocity component, time change component) and processing them through separate convolution filtering operations, the system can extract distinctive features from each segment and combine them for more accurate authenticity verification.
2Device complexity
If traditional radar data recognition methods are used, then the system is simpler, but the ability to distinguish between real and fake objects deteriorates
Solution Approach 1:
The patent introduces convolution filtering as an intermediary processing step between the radar data input and the recognition model. This intermediary layer extracts and emphasizes distinctive spatial features and temporal patterns, acting as a mediator that transforms raw radar data into enhanced feature representations that are more reliable for distinguishing real objects from fake objects.
Solution Approach 2:
The patent changes the parameter configuration of the input data by creating multiple dimension configurations (angle dimension, range-velocity dimension, time frame dimension). These parameter transformations reveal hidden patterns and spatial relationships that are not apparent in the original data, enabling more reliable authenticity verification while maintaining reasonable system complexity.
Data Source
AI summary
A processor-implemented radar data recognition method including generating a plurality of pieces of input data, with respectively different dimension configurations, based on radar data of an object; and outputting a recognition result of the object based on the generated plurality of pieces of input data using a recognition model.