Radar Data Classification via Logarithmic Polar Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radar data classification methods face challenges in accurately identifying objects in varying distances and angles, particularly for infrequently occurring traffic-relevant objects, due to the need for extensive training data and increased computing time.
Innovation Solution
Transforming radar data into logarithmic polar coordinates, using a fixed point as the origin, allows the classifier to separate changes in object size and orientation from changes in perspective, reducing the requirement for diverse training data and enabling efficient classification of objects regardless of scale and rotation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If radar data are classified using conventional methods without coordinate transformation, then the classifier must handle multiple variations of objects at different distances and angles, but this requires extensive training data and increases computing time
Solution Approach 1:
The patent transforms radar data from Cartesian coordinates to logarithmic polar coordinates, adding a dimensional transformation that separates perspective changes from object variations. This coordinate system change allows the classifier to handle objects at different distances and angles more effectively, improving classification reliability while reducing the need for extensive training data and computing time
2Productivity
If radar data are transformed into logarithmic polar coordinates, then the classifier can separate perspective changes from object variations, but this requires additional processing steps
Solution Approach 1:
The patent applies parameter changes by transforming the coordinate system from Cartesian to logarithmic polar coordinates. This parameter transformation fundamentally changes how the data is represented, allowing the classifier to distinguish between objects at different distances and angles more effectively, thereby improving classification efficiency despite the additional processing step
3Measurement precision
If the classifier is trained on diverse training data to handle objects at various distances and angles, then classification accuracy improves, but the amount of training data required increases significantly
Solution Approach 1:
By introducing the logarithmic polar coordinate transformation, the patent creates a new dimensional representation that inherently handles variations in distance and angle. This transformation allows the classifier to achieve high identification accuracy for objects at various positions without requiring proportional increases in training data quantity, as the coordinate transformation itself captures the geometric variations
Data Source
AI summary
A method for classifying radar data, which have been obtained by registering radar radiation emitted from a transmitter and reflected by at least one object using at least one detector. The method includes: providing radar data, which include observations of a setting recorded at different points in time; ascertaining at least one portion of the radar data, which is rotated and/or scaled in at least one of the observations as compared to at least one other of the observations; ascertaining a fixed point of the rotation and/or scaling; transforming at least one two-dimensional representation of at least one part of the observations into logarithmic polar coordinates using the ascertained fixed point as the origin; and mapping the at least one transformed two-dimensional representation onto at least one class of a predefined classification via at least one classifier, which encompasses a neural network that includes at least one convolution layer.