Radar Target Detection by Sub-Area CNNs for Low False Alarms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing target detection methods face challenges in balancing false alarm rates and missing targets due to threshold settings in CFAR detection, and AI-based methods lack real-time performance and practicality due to large detection model scales and computation requirements.
Innovation Solution
A target detection method utilizing a radar system with a control device that processes sensing data through convolutional neural networks to provide real-time indication of targets in sub-areas, reducing false alarms and improving accuracy without converting data into 3D images or performing extensive computations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CFAR detection with threshold settings is used, then detection sensitivity is improved, but false alarm rate increases
Solution Approach 1:
The patent divides the sensing data into multiple sub-areas and processes each sub-area independently through the neural network. This segmentation allows the system to provide indication information for each sub-area, reducing the impact of false alarms in one area on the overall detection system while maintaining high detection sensitivity through adaptive thresholding in each segment.
Solution Approach 2:
The patent implements a feedback mechanism where the neural network learns from training samples with labels indicating target presence in sub-areas. The model adjusts its detection thresholds and parameters based on feedback from training data, optimizing the balance between detection sensitivity and false alarm rate through continuous learning rather than static threshold settings.
2Measurement precision
If large-scale AI detection models are used, then detection accuracy is improved, but real-time performance deteriorates
Solution Approach 1:
The patent segments the detection task into processing multiple sub-areas independently. By dividing the sensing data into smaller sub-areas and processing them through the neural network separately, the system reduces the computational burden on each processing unit while maintaining overall detection accuracy through the collective results from all sub-areas.
Solution Approach 2:
The patent applies partial action by focusing the neural network processing on essential features and key sub-areas rather than processing all data with equal depth. The model learns to identify critical patterns in training data and applies this knowledge to make accurate detection decisions with reduced computational effort, achieving sufficient accuracy without exhaustive processing.
Data Source
AI summary
Provided in the present disclosure are a target detection method and apparatus, and a target detection model training method and apparatus. The target detection method includes: acquiring sensing data of an area to be sensed; and obtaining a detection result according to the sensing data and a target detection model, where the detection result includes a plurality of pieces of indication information, each piece of indication information of the plurality of pieces of indication information corresponds to a sub-area in the area to be sensed, and the each piece of indication information is used to indicate whether there is a target in the sub-area corresponding to the each piece of indication information.