Food waste in airlines poses a significant threat to sustainability, leading to wasted resources, financial losses, and environmental harm. Traditional quantification methods often rely heavily on manual data collection or subjective visual assessments, which introduces inefficiencies and inaccuracies. This research proposes a novel framework leveraging image recognition techniques and deep learning algorithms on a cloud-based platform. Using Google Vertex AI, an AI system was developed to analyze images of leftover food on airline trays, achieving a remarkable accuracy of 92.5%. This result was achieved by employing a dropout rate of 0.3, image augmentation, and the ReLU activation function. The findings demonstrate that the proposed algorithm is robust in recognizing and estimating food leftovers. By enabling accurate food waste quantification, this research empowers airlines to develop data-driven waste reduction strategies, contributing to a more sustainable and efficient food service system in the aviation industry. Ultimately, this solution enhances sustainability, achieves significant cost savings, and improves operational efficiencies across the industry.