Dynamic eye tracking in 3D spaces provides a promising method for understanding human perception and judgment of landscape qualities. Studies applying this technology to outdoor landscapes remain scarce. This experimental study combines outdoor eye-tracking technology with artificial intelligence (computer vision) to represent, visualize, and interpret human observational patterns in a rehabilitated agricultural landscape. It examines what individuals observe during free-viewing of an outdoor environment and how these observations relate to landscape elements. Findings reveal consistent patterns (interest in mid-ground, water features) that align with established landscape theories and highlights phenomena requiring further investigation to enhance the understanding of the relationship between design theories and observational behavior.