Abstract: With the continuous development of technology, the detection of high-speed and highly maneuverable weak targets has become a current hot topic. Although signal accumulation can improve the echo signal-to-noise ratio, the high maneuverability of the target will cause range migration and Doppler frequency migration in the echo signal, affecting the effective coherent accumulation of the signal. This paper proposes a joint algorithm based on time reversal transformation - second-order Keystone transformation - fractional Fourier transform and improved Radon transform. The algorithm first eliminates the first-order range migration through time reversal transformation, then corrects the second-order range migration through SKT, estimates the acceleration by performing FRFT on the target range cell, and constructs a phase compensation function to correct the DFM of the pulse compression signal. Then, it performs an improved Radon transform to estimate the target velocity and construct a velocity phase compensation function. Finally, it performs a fast Fourier transform in the slow time dimension to complete the coherent accumulation. Simulation experiments prove that this algorithm can eliminate the influence of range migration and Doppler frequency migration, does not require a complex parameter search process, and solves the problem of velocity loss in the TRT process. Compared with the comparison algorithm in the paper, it has the characteristics of low complexity and achieves a good balance between complexity and detection performance.