For the nondestructive detection of small-scale damage in large metallic isotropic thin-plate structures, existing ultrasonic guided wave damage detection technologies have achieved basic damage localization. However, due to limitations in detection principles, imaging algorithms, and hardware systems, issues such as low positioning accuracy, low detection efficiency, and the inability to acquire morphological information persist. This study investigates a finite-difference-based reverse time migration (RTM) imaging algorithm, grounded in the complex interaction mechanisms between laser ultrasonic guided waves and structural damage. The algorithm employs two-dimensional time-domain acoustic wave finite-difference numerical calculations to compute both the incident and back-propagated scattered wavefields, thereby improving computational efficiency. Signal preprocessing and calibration methods for array wavefields are studied to maximize the extraction of scattered wavefields. A Gabor wavelet is used as the excitation signal, and a preprocessing scheme is proposed to match simulated and acquired array wavefields, including excitation time calibration, amplitude calibration, and velocity calibration. Based on finite element simulations and experimental laser ultrasonic data, array wavefield data of irregular damages with indentations and sharp corners are obtained. The damage is reconstructed using the FD-RTM imaging method. The research results provide guidance for the rapid reconstruction method of damage morphology.