CoolFace
Apppublic

jerrychen76/fruit-ripeness-recognition

sourceHugging Facemitupdated 7mo agoView on Hugging Face
0likes
App README

Fruit Ripeness Recognition

This Space is a fruit ripeness recognition demo built with a trained YOLO model. Users can upload an image, and the system will detect the fruit and predict its ripeness level with annotated output.

Overview

This project recognizes 5 types of fruits, and each fruit is divided into 3 ripeness levels:

  • —Unripe
  • —Ripe
  • —Overripe

That gives a total of 15 classes.

Fruit Categories

The model supports the following fruits:

  • —Orange
  • —Apple
  • —Mango
  • —Banana
  • —Tomato

Classes

The 15 classes in this project are:

  • —orange_unripe
  • —orange_ripe
  • —orange_overripe
  • —apple_unripe
  • —apple_ripe
  • —apple_overripe
  • —mango_unripe
  • —mango_ripe
  • —mango_overripe
  • —banana_unripe
  • —banana_ripe
  • —banana_overripe
  • —tomato_unripe
  • —tomato_ripe
  • —tomato_overripe

Features

  • —Upload an input image
  • —Detect fruits in the image
  • —Predict ripeness level for each detected fruit
  • —Return an annotated output image with bounding boxes and labels
  • —Display performance information such as inference time and FPS

Input

  • —One fruit image uploaded by the user

Output

  • —Annotated image with detection boxes
  • —Predicted fruit type
  • —Predicted ripeness level
  • —Inference time
  • —Total processing time
  • —FPS

Model

This demo uses a trained YOLO model stored as:

  • —best.pt

Project Files

  • —app.py — Gradio application for inference
  • —requirements.txt — required Python packages
  • —best.pt — trained model weights
  • —README.md — project description for Hugging Face Space

Example Tasks

This model can help with tasks such as:

  • —Fruit ripeness inspection
  • —Agricultural product monitoring
  • —Smart food sorting
  • —Computer vision demonstration for ripeness classification

Notes

  • —This Space runs on Hugging Face Spaces.
  • —Performance may be slower on free CPU hardware.
  • —The displayed FPS depends on the current deployment environment.

Usage

  1. 1.Upload a fruit image.
  2. 2.Wait for the model to process the image.
  3. 3.View the annotated result with fruit type and ripeness level.