Multi-target constant false alarm rate detection method based on deep neural network

Abstract

The invention discloses a multi-target constant false alarm rate detection method based on a deep neural network, and the method comprises the steps: training a pre-detector based on the deep neural network through building a simulation data set employing a data enhancement technology, and carrying out the classification of a radar signal peak value, so as to distinguish whether the radar signal is a target or a clutter; removing the target detected by the pre-detector from the original background sample to form a reduced sample; performing background level estimation by using an approximate maximum likelihood estimator based on Taylor series based on the reduced sample to obtain a false alarm adjustment threshold, removing a target lower than the threshold in a pre-detection result, and outputting a final detection result. The method provided by the invention does not need to depend on a pre-estimated background level to detect the target, and can still maintain excellent detection performance in a scene with very high target density.

Description

technical field [0001] The invention belongs to the technical field of frequency modulated continuous wave (Frequency Modulated Continuous Wave, FMCW) radar multi-target constant false alarm rate (Constant False Alarm Rate, hereinafter referred to as CFAR) detection, and specifically relates to a multi-target constant false alarm rate based on a deep neural network Detection method. Background technique [0002] Multi-object detection is very challenging, especially in scenes with densely distributed objects. In traditional CFAR detection methods, the detection threshold is determined based on pre-estimated background levels. However, interfering targets will inevitably lead to inaccurate background level estimation, resulting in poor detection performance. Contents of the invention [0003] Aiming at the shortcomings of the traditional CFAR detection method in multi-target scene detection performance degradation, the present invention proposes a multi-target constant fa...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the shortcomings of the traditional CFAR detection method in multi-target scene detection performance degradation, the present invention proposes a multi-target constant false alarm rate detection method based on a deep neural network, which converts target detection into radar peak sequence classification through a deep neural network detector problem to improve detection performance without relying on background level estimates

Images