To solve the problem of detecting the space-borne radar linear frequency modulated (LFM) pulse signal contaminated by spurious clutter and interference under a low signal-to-noise ratio (SNR), a detection and parameter estimation algorithm based on the time-frequency image enhancement and Hough transform (HT) is proposed. First, short-time Fourier transform (STFT) is carried out on the space-borne radar LFM pulse signal with the Gaussian window to acquire the time-frequency spectrum; the spectrum is then converted into the time-frequency image. Second, in order to observe the weak signal’s details from the image, contrast stretching is implemented on the time-frequency image and the trailing induced by the spurious clutter is eliminated to strengthen the chirp line of the LFM signal. Third, the line is detected through HT on the enhanced time-frequency image, and the coordinates and gray values of the pixels passed through by the line are extracted in the time-frequency image; then these gray values are filtered by the median filter before being binarized with the gray threshold; the anti-pulse-splitting mechanism is adopted to determine the start and the end of the chirp line segment. Measured data experiments show that the method can effectively detect the space-borne radar LFM pulse signal under a low SNR environment, determine the pulse’s arrival time and end time, estimate its parameters, and is superior to direct STFT-Radon transform and direct STFT-HT.