<p>Accurate estimation and classification of impact energy levels in composite materials remain challenging due to the complex and barely visible nature of impact-induced damage. To address this issue, this study proposes a novel framework that integrates ultrasonic imaging with artificial intelligence for intelligent impact energy assessment. Composite specimens were subjected to controlled impact tests, and internal damage was characterized using both B-scan and C-scan ultrasonic modalities acquired with the UPKT36 system. A set of five hybrid health indicators was then developed by fusing features extracted from the front and back surfaces of each specimen. Three indicators were derived from segmented subsignals, while the remaining two quantified the damage area extracted from the C-scan images and the mean pixel intensity extracted from the B-scan images of both the front and back surfaces, respectively. These indicators capture complementary geometric and statistical characteristics of impact-induced damage, including damage width, height, eccentricity, and intensity-based descriptors. The extracted features were used as inputs to an Adaptive Neuro-Fuzzy Inference System (ANFIS) model to predict the corresponding impact energy levels. Experimental results demonstrate that the fusion of multi-view ultrasonic information with the proposed hybrid indicators significantly improves prediction accuracy and robustness, confirming the effectiveness of the proposed methodology for data-driven damage assessment in composite structural health monitoring.</p>