https://github.com/srbhr/fruits_360

Fruits Detection using CNN.
https://github.com/srbhr/fruits_360

Last synced: 12 days ago
JSON representation

Fruits Detection using CNN.

• Host: GitHub
• URL: https://github.com/srbhr/fruits_360
• Owner: srbhr
• License: gpl-3.0
• Created: 2019-10-26T09:41:04.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2024-08-30T23:51:43.000Z (2 months ago)
• Last Synced: 2024-10-03T12:42:18.296Z (about 1 month ago)
• Language: Jupyter Notebook
• Size: 11.6 MB
• Stars: 32
• Watchers: 5
• Forks: 13
• Open Issues: 7
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Fruits_360

Fruits Detection using CNN.

Dataset used :

### Fruits 360

 A dataset of images containing fruits and vegetables

##### Dataset properties

- Total number of images: 82213.

- Training set size: 61488 images (one fruit or vegetable per image).

- Test set size: 20622 images (one fruit or vegetable per image).

- Multi-fruits set size: 103 images (more than one fruit (or fruit class) per image)

- Number of classes: 120 (fruits and vegetables).

- Image size: 100x100 pixels.

The Dataset Can be found over :  https://www.kaggle.com/moltean/fruits  and  https://github.com/Horea94/Fruit-Images-Dataset 

This is the work  of  Horea Muresan, [Mihai Oltean](https://mihaioltean.github.io/), [Fruit recognition from images using deep learning](https://www.researchgate.net/publication/321475443_Fruit_recognition_from_images_using_deep_learning), Acta Univ. Sapientiae, Informatica Vol. 10, Issue 1, pp. 26-42, 2018.

The paper introduces the dataset and an implementation of a Neural Network trained to recognized the fruits in the dataset.

## How to use this

### Requirements

The requirements.txt file, has all the packages that were in the environment at the time of training. 

* Tensorflow 2.0  (Tensorflow-GPU was used)

* Keras 2.3.1

* Matplotlib

* Numpy

### Usage 

The Images to be predicted are put under the fruits/test_images folder.

This model is pretrained with and weights is a H5py file.  Named 'Fruits_360.h5'.

The fruits.py file contains the Network Model and was used to train it.