The study ‘Kitchen Food Waste Image Segmentation and Classification for Compost Nutrients Estimation’ examines the potential of artificial intelligence to quantify the nutritional properties of compost made from daily food waste. The LILA home composter concept relies on image segmentation and classification techniques to analyze high-resolution images of food waste. With a dataset annotated with segmentation masks of nutrition-rich categories, this research explores the efficacy of semantic segmentation models in determining compost quality.

• A comprehensive image dataset for kitchen food waste with 19 nutritional categories.
• Evaluation of various semantic segmentation models to assess compost quality.
• SegFormer model achieves the highest mean Intersection over Union (mIoU) score.
• Potential application in the accurate estimation of Nitrogen, Phosphorus, or Potassium in compost.

This paper underscores the innovative ways in which AI can be used to support sustainability initiatives, especially in waste management and recycling. By improving the understanding of the composition and value of compost, this technology could help individuals and communities make more informed decisions about waste management and promote environmentally-friendly practices. Read More