Produce Quality Detection

This project applies computer vision to automated produce quality detection and was built during an AI research internship with Youth in AI. The work covered the pipeline around the model as well as the model itself: preparing a custom produce dataset, annotating examples, preprocessing inputs, and fine-tuning YOLOv8 for the detection task.

Python and OpenCV support the image-processing workflow around YOLOv8. A substantial part of the engineering work involved making the annotation and preprocessing process more consistent so that training had a more dependable input. The result was an end-to-end detection pipeline, connecting dataset preparation to inference rather than treating a trained model as the complete deliverable.