Automated pavement crack detection is essential for maintaining road infrastructure, yet conventional methods often face challenges such as data scarcity and reliance on manual inspection. This study integrates synthetic data generation with deep learning to address these limitations and improve the accuracy of automated crack detection. Using the Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) to generate realistic pavement crack images, the U-Net segmentation model is optimized. The U-Net model, trained with an augmented dataset containing 1000 synthetic images (400 % increase), achieved superior performance, with a dice coefficient of 0.961 and an intersection over union (IoU) of 0.925. Its generalization capabilities were validated across multiple real-world datasets, demonstrating reliable performance under diverse crack patterns and imaging conditions. By addressing the challenges of data scarcity and improving segmentation accuracy, this research offers a scalable and efficient framework for automated pavement evaluation. The findings underscore the transformative potential of generative models in advancing automation in construction, reducing manual labor, and enabling data-driven solutions for infrastructure management.