Download PDFOpen PDF in browser

Fruits Fresh and Rotten Detection Using CNN & Transfer Learning

EasyChair Preprint no. 9705

8 pagesDate: February 14, 2023

Abstract

The economic development of our nation is significantly influenced by agriculture. Fruit production with high yields and productive growth are crucial to the agriculture sector. 30 to 50 percent of the gathered fruit is wasted because there aren't enough qualified workers. Additionally, fruit identification, classification, and grading are not done accurately due to human perception subjectivity. Therefore, the fruit sector must impose an automation system. In order to save labour, production costs, and production time, this research suggests a method based on recognizing fruit flaws in the agriculture sector. These flawed fruits can infect healthy fruits if we are unaware of them. As a result, we suggested a methodology to stop corruption from spreading. From the input fruit photos, the suggested model distinguishes between fresh and rotting fruits. Apples, bananas, and oranges are the three types of fruits I used in this project. Softmax is used to categories the input fruit image into fresh and rotten fruits, and a convolutional neural network (CNN) is utilized to extract features from the input fruit image. Utilizing the Kaggle dataset, the suggested model's performance is assessed. This results in 81% accuracy. The findings demonstrate that the suggested CNN model can successfully classify both fresh and rotting apples. The suggested study investigated strategies for classifying fresh and rotting fruits using transfer learning models. The suggested CNN model outperforms transfer learning models and prior art methods in terms of performance.

Keyphrases: Agriculture, CNN, Transfer Learning Models

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:9705,
  author = {G Anitha and P Thiruvannamalai Shivashankar},
  title = {Fruits Fresh and Rotten Detection Using CNN & Transfer Learning},
  howpublished = {EasyChair Preprint no. 9705},

  year = {EasyChair, 2023}}
Download PDFOpen PDF in browser