Sea-surface small target detection based on PCA and weighted KNN algorithm

The method based on Principal Component Analysis (PCA) and weighted k-nearest neighbor classifier (KNN) is proposed to solve the problem for the traditional KNN algorithm not fulfilled the false alarm probability requirements of the radar targets detection. The PCA method is used to calculate the feature contribution rate of features used to distinguish sea clutter and target echo. A weighted KNN detector is proposed to improve the detection performance on the basis of ensuring the false alarm probability by the feature contribution rate for distance weighted. Finally, the measured datas of the IPIX radar are used for verification. The results show that when the observation time is 0.512 seconds, the detection probability is 13.27% higher than other detectors, and the false alarm rate can be better control detection probability. Therefore, the detector proposed can effectively improve the detection performance of the sea surface small targets.