Automatic Radar Spectrum Labeling Using Reference Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radar processing systems require numerous in-flight measurement campaigns and extensive human intervention to create a reference database for training artificial intelligence algorithms to identify ground surfaces, which is inefficient and labor-intensive.
Innovation Solution
A method for automatic labeling of radar spectra using a reference system, such as a reference sensor or digital terrain model, to associate true identifiers with radar data, enabling the creation of a training database with minimal human intervention and reducing the number of required measurement campaigns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If numerous flight measurement campaigns are conducted to collect spectra for training database, then the quality and quantity of training data is improved, but the time and resources required are excessive
Solution Approach 1:
The patent uses a reference system (such as optical sensors, LIDAR, or digital terrain models) to create copies or proxies of ground surface information that can be directly associated with radar spectra. Instead of manually collecting and labeling data through extensive flight campaigns, the reference system automatically provides ground truth labels by capturing images or terrain data of the illuminated surface, which are then linked to corresponding radar spectra to build the training database efficiently
Solution Approach 2:
The reference system acts as an intermediary between the radar system and the ground surface. It captures information about the illuminated surface (through optical imaging, LIDAR, or terrain models) and provides this information as labels for the radar spectra. This intermediary approach automates the labeling process and eliminates the need for manual expert analysis during flight campaigns
2Measurement precision
If expert personnel manually analyze and label spectra, then the accuracy of ground surface identification is improved, but the labor and time required are excessive
Solution Approach 1:
The system enables self-service labeling by automatically associating reference system data with radar spectra without requiring manual expert intervention. The reference system (optical sensors, LIDAR, or digital terrain models) automatically captures ground surface information and links it to the corresponding radar spectra, creating labeled training data autonomously. This eliminates the need for expert personnel to manually analyze and label each spectrum while maintaining high accuracy
3Productivity
If the reference database is built with minimal human intervention, then the efficiency of database creation is improved, but the accuracy of labels may be compromised
Solution Approach 1:
The reference system serves as a reliable intermediary that automatically captures ground surface information with high accuracy. By using established sensing technologies (optical cameras, LIDAR) or authoritative data sources (digital terrain models), the system ensures that the automatic labeling process maintains high precision without requiring human intervention. The reference system's measurements are trusted to provide accurate ground truth labels
Solution Approach 2:
The patent replaces the mechanical process of manual expert analysis with automated sensing and data processing systems. Instead of experts visually inspecting and labeling spectra, the reference system automatically captures ground surface data and associates it with radar spectra through computational processes, achieving both high efficiency and maintained accuracy through technological substitution
Data Source
Figure 1
Figure 2
AI summary
The present invention relates to a computer-implemented method (100) for the automatic labeling of data corresponding to a radar spectrum, the radar spectrum being acquired by means of a radar on board an aircraft and illuminating a ground surface, in order to train an artificial intelligence algorithm for identifying the ground surface, the method consisting of: acquiring (110) a radar spectrum of a domain observed by the radar; obtaining (120, 140), by means of a reference system, a true identifier of a ground surface illuminated by the radar during the acquisition of the radar spectrum; and associating (150), in a labeled data, the data corresponding to the radar spectrum and the true identifier obtained.