针对超高射速弹丸出膛时刻难以精确获取的问题,提出了一种图像边缘检测的方法来获取出膛时刻。采用连续波雷达照射炮口,在每发弹丸的出膛时刻,多普勒回波信号产生突变。通过短时傅里叶变换和时频域最小噪声谱估计降噪处理后,利用图像边缘检测在时频域对回波信号突变点进行检测和定位,找出每发弹丸回波的突变点作为弹丸的出膛时刻。通过对实测信号处理验证了本算法的正确性和可行性。

In view of difficulty on accurately obtain fire artillery projectile's muzzle-leaving time, a method based on image edge detection was presented in the paper. Continuous wave radar was used to beat down muzzle, the echo signal generated a mutation at each projectile muzzle moment. The mutation point can be detected and located after STFT, the noise reduction by minimum statistics noise estimation and image edge detection in time-frequency domain. It is verified correct and feasible by the processing of the measured signal. It provides reference for obtaining the ultra high rate of fire artillery projectile's muzzle-leaving time.