It is a challenge to accurately extract a traveling wavefront because of the high resistance grounding fault, zero-crossing fault and high frequency noise, resulting in low reliability of traveling wave-based protection and the fault location method. Since the traveling wave has full time-frequency characteristics, the real-time detection of traveling wave in the time-frequency domain will contain panoramic fault information, which can realize the observability of the fault characteristics. Fusing the traveling wave in the time-frequency domain, a new traveling wave representation in 3D subspace, called full waveform, is proposed. A novel time-frequency analysis method based on variational mode decomposition (VMD) and Wigner-Ville distribution (WVD) is presented. The traveling wave signals are decomposed by VMD, and the mode components are extracted. Wigner-Ville distributions of mode components are computed and the Wigner-Ville of each component signal is added linearly to reconstruct the Wigner-Ville of the original signal. The proposed method is applied to analyze the simulated and practical measured traveling wave signals. The results show that the proposed method can guarantee good noise suppression of the VMD, and the high time-frequency resolution and energy aggregation of the Wigner-Ville distribution. The proposed method produces a true and accurate extraction of the full waveform. Applied to a 1:1 real model experimental platform, it can realize accurate fault location in a way that is not affected by high resistance grounding fault, zero-crossing fault, noise interference and sampling rate. It is expected to greatly improve the reliability of the traveling-wave-based protection and fault location method. This work is supported by the National Natural Science Foundation of China (No. 52077008).