Target length estimation based on quadratic statistical CFAR processing

Estimating the radial size of targets using radar high resolution range profile (HRRP) can provide important feature criteria for target classification and recognition. In practice, the bi-directional constant false alarm rate (CFAR) threshold method is often used to discriminate the signal support area and noise area of targets. Under the condition of low signal-to-noise ratio, the size estimation accuracy of existing methods is low, and it is easy to miss detection when the edge signal of target support area is weak. In order to solve the above problems, this paper judges the target-noise boundary by multiple detections with low threshold, and further improves the size estimation accuracy by edge optimization after searching the target-noise boundary. Experimental verification and comparative analysis show that the average estimation error of this method is less than 10%, which is significantly lower than the existing CFAR threshold size estimation method, and can effectively avoid missing detection of edge weaknesses and strong noise interference.