The Kalman filter algorithms can not solve the problem of the time-frequency localization, and it often switches between the low-frequency filter and high-frequency filter by the mobile detector (The changeable dimension filter algorithms) or Markovian transfer probability matrix (VD algorithms), so it has delay and some influence of the the mobile detector and Markovian transfer probability matrix. Hence, the filter error of the Kalman filter algorithms is obviously big. This paper brings the new mixed filter algorithms (WK algorithms) of the 2D Wavelet transform and Kalman Filter, which has the characteristics of the well time-frequency localization and real-time. The WK algorithms is used to process the radar information and makes the filter estimated data to approach the true track.