Resonance between wind turbine towers and rotors can severely compromise the safety of wind turbines. This study proposes a measurement method based on the digital image correlation (DIC) technique for determining the natural frequencies of prestressed concrete towers. The approach employs high-pass filtering and wavelet transform to reconstruct vibration signals, followed by the use of Welch’s method to calculate the power spectrum of the reconstructed signals. The natural frequencies are identified from the spectral peaks, and the first- and second-order modes of the tower are determined through singular value decomposition (SVD). The measured first-order natural frequencies of the concrete tower under three operating conditions—turbine during stoppage, stable operation, and transition from stoppage to stable operation—are 0.2631 Hz, 0.2636 Hz, and 0.2782 Hz, respectively, whereas the second-order natural frequencies are 0.8624 Hz, 1.0397 Hz, and 0.8374 Hz, respectively. The first-order mode shapes correspond to unidirectional bending of the tower, and the second-order mode shapes correspond to reverse bending. The results demonstrate that the DIC method is an effective, convenient, and safe approach for measuring the natural frequencies of wind turbine towers, with significant practical value.