The use of fiber-reinforced composite materials in marine applications is limited by uncertainty surrounding their long-term fatigue behavior and micro-damage tolerance. This thesis aims to present and validate an experimental framework to detect and classify mechanical micro-damage in unidirectional carbon-fiber composites using acoustic emission (AE) monitoring and X-ray micro-computed tomography (micro-CT). AE monitoring provides real-time insight into the evolution of internal damage by capturing elastic waves emitted during micro-structural failure events, while micro-CT offers high-resolution visualization of internal damage states before and after mechanical loading. A comprehensive analysis was conducted involving signal processing, (normalized) frequency spectrum characterization, and unsupervised machine learning to classify AE events by damage type. This classification was subsequently validated against micro-CT scans. Results challenge the common assumption that AE signals with dominant low-frequency contributions are reliably indicative of matrix cracking. The proposed AE framework, when validated with micro-CT, shows promise for enabling accurate in-situ damage monitoring of composite structures in offshore environments. This approach supports the broader adoption of composites by improving confidence and knowledge about their structural integrity over time.