Concerning the overlap of ground penetrating radar (GPR) signal reflections at the different asphalt layers, which reduces the measurement accuracy of asphalt surface thickness, the WRELAX algorithm for time-delay estimation was applied to analyze the air-launched GPR signal reflections with 1GHz center frequency to improve the detection accuracy of thin layers of asphalt pavement in this paper. The results were verified by simulation experiment and field test. The research results show that the WRELAX algorithm detects iteratively strong reflections and then subtracted from the original signal to reveal the weak reflections. Time-delay and amplitude can be calculated by the WRELAX algorithm. The WRELAX algorithm in 1GHz GPR signal processing can effectively detect thin layers of asphalt pavement, and variation ranges of estimation errors are relatively small. The algorithm is robust and its effect is better than 2GHz GPR signal processing. The maximum relative estimation error of thin layer estimated by the WRELAX algorithm is less than 5%, which meets the requirements for detection and evaluation of asphalt thin layers.