Smart Sorting System
This project focuses on using transfer learning techniques to identify rotten fruits and vegetables. By
automating the sorting process, it aims to improve efficiency and accuracy in the agricultural and food
industry.
Smart Sorting Transfer Learning for
Identifying Rotten Fruits and Vegetables
Branch Name Track Team Lead Team Members
Computer Science and Artificial Intelligence Thapeta Gangadevi A Anitha, Udayagiri
Engineering and Machine Learning Anjali, Thota
Dakshitha Raja
Project Scenarios
Fresh Shipment Scanning
Ensures only fresh produce
2 reaches customers, minimizing
Automating Sorting waste.
Reduces human error and 1
labor intensity, enhancing
efficiency.
Smart Home Monitoring
3
Alerts users about items nearing
spoilage, thus reducing food
waste.
Architecture and Prerequisites
Architecture Prerequisites
The system integrates a model with a Flask Users must install Anaconda and several Python
application for image analysis. packages, including TensorFlow and Flask.
Project Objectives
Understand Deep Data Preprocessing Develop Data
Learning Knowledge Visualization Skills
Grasp fundamental concepts Learn data transformation Enhance abilities for insights
and techniques. techniques. analysis.
Project Flow
1 2 3
User Image Selection Model Processing Results Display
The model processes the
The user selects an image The results are displayed on the
selected image and predicts the
through the UI. UI for the user to see.
outcome.
Data Collection and
Preparation
1 Dataset
The project utilizes a dataset of 28 classes of fruits and
vegetables sourced from Kaggle.
2 Data Visualization
Techniques are employed to visualize and analyze the
dataset for better understanding.
3 Data Augmentation
Various transformations are applied to increase the dataset's
diversity, aiding model robustness.
Application Development
1 HTML Pages
Three HTML files for user interaction.
2 Flask Integration
Connects the UI to the predictive model.
3 Model Testing
Tests the model's accuracy in identifying produce.
Advancements in Smart Sorting System
The Smart Sorting system demonstrates
significant advancements in fruit and vegetable Efficiency Enhancement
sorting through AI and transfer learning. By Automates sorting processes to improve
automating processes, it enhances efficiency, efficiency.
reduces waste, and improves quality control in the
agricultural sector. Future work can expand the Waste Reduction
model's capabilities and applications.
Significantly reduces waste in the agricultural
sector.
Quality Control Improvement
Enhances quality control measures in sorting.
CONCLUSION:
Smart Sorting using transfer learning effectively
automates the detection of rotten fruits and vegetables,
improving accuracy and reducing manual effort. By
leveraging pre-trained models, it ensures faster, scalable,
and reliable quality control—helping reduce food waste
and enhance supply chain efficiency.
