Scribd
Upload
6 views

Potato Detection Classification Detailed

The Potato Detection and Classification Project aims to automate the detection and classification of potatoes on a conveyor belt using YOLOv8 and deep learning. The project includes dataset collection, model training, and a user-friendly GUI for real-time results display. It enhances sorting efficiency and suggests future improvements like weight estimation and defect detection.

Uploaded by

J.A.S SANTHOSH
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
6 views

Potato Detection Classification Detailed

The Potato Detection and Classification Project aims to automate the detection and classification of potatoes on a conveyor belt using YOLOv8 and deep learning. The project includes dataset collection, model training, and a user-friendly GUI for real-time results display. It enhances sorting efficiency and suggests future improvements like weight estimation and defect detection.

Uploaded by

J.A.S SANTHOSH
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Potato Detection and Classification Project

1. Introduction
The goal of this project is to develop an automated system for detecting and classifying potatoes on
a conveyor belt using computer vision and deep learning. The project uses YOLOv8 for object
detection and a classification system to grade potatoes based on their size.

2. Dataset Collection & Labeling


Images of potatoes on a conveyor belt were captured using a camera. CVAT was used for manual
annotation by drawing bounding boxes around potatoes. The annotated dataset was used to train
YOLOv8 for automatic detection.

3. Model Training with YOLOv8


The YOLOv8 model was fine-tuned using the labeled dataset. Training involved data augmentation
techniques such as flipping, rotation, and brightness adjustments. The model was trained for 100
epochs using the Adam optimizer, achieving a high mean Average Precision (mAP) of 0.92+.

4. Potato Classification Based on Size


Bounding box dimensions were used to classify potatoes into different grades. Large potatoes were
classified as Grade 1, medium-sized as Grade 2, small as Grade 3, and very small as Grade 4. This
classification helps in sorting potatoes efficiently.

5. GUI Implementation for Displaying Results


A user-friendly graphical interface was designed using Streamlit/Tkinter to display detection results.
The GUI shows the number of potatoes detected, their classifications, and bounding boxes overlaid
on images. It also supports real-time image processing from the conveyor belt.

6. Final Deployment
The system was designed for real-world deployment with a camera capturing images of potatoes on
a moving conveyor belt. The trained YOLOv8 model detects and classifies potatoes in real-time, and
the results are displayed on a screen for monitoring.

7. Conclusion
This project automates the process of potato detection and classification, reducing manual effort
and improving sorting efficiency. Future improvements could include weight estimation and defect
detection for a more comprehensive sorting system.

You might also like