In the detection of the high-speed and weak targets, Radon-Fourier Transform(RFT) may encounter the problems of heavy computational quantity and real-time processing. To solve the problems, the Frequency Shear Radon-Fourier Transform(FSRFT) is proposed, which simplifies process and decreases computation effectively under the same integration gain and detection performance, and is beneficial to sliding window process and parallel computing for radar tracking as well. The experiment results show that the detection performance can almost achieve the theoretical value under ideal condition, and the parallel processing capability is lower by one order of magnitude than the existing fast methods.