Smart Quality Testing System for Flipkart GRID 6.0

This repository contains the code and resources for a smart quality testing system developed as part of the Flipkart GRID 6.0 competition. The project leverages computer vision and natural language processing to automate quality checks using camera vision technology without the need for extra hardware. It includes three main components: Freshness Detection, QC Process Automation, and a PaddleOCR Server for text extraction.

Project Overview

The goal of this project is to create a cost-effective, intelligent system capable of analyzing product quality, extracting text information from packaging, and matching brand details, all with the use of a single camera.

---

How to Run the Project

Application 1: Freshness Detection

Follow these steps to set up and run the Freshness Detection module.

---

Step 1: Clone the Repository

git clone https://github.com/Gopinathv19/FlipKartGrid6.0.git
cd FreshnessDetection

---

Step 2: Set Up the Virtual Environment and Install Requirements

Create a virtual environment (optional but recommended) and install all required dependencies:

# Create a virtual environment (optional)
python -m venv venv
source venv/bin/activate    # On Linux/MacOS
venv\Scripts\activate       # On Windows

# Install dependencies
pip install -r requirements.txt

---

Step 3: Set Up the MySQL Database

Set up a MySQL database and create the table with the schema below:

CREATE TABLE FreshnessData (
    Sl_no INT AUTO_INCREMENT PRIMARY KEY,
    Timestamp DATETIME NOT NULL,
    Produce VARCHAR(50) NOT NULL,
    Freshness VARCHAR(20) NOT NULL,
    Expected_Life_Span_Days INT NOT NULL
);

Ensure your database connection details (username, password, host, and database name) are configured correctly in the code.

---

Step 4: Run the Application

Run the application using Streamlit from the command line:

streamlit run bananaFreshnessDetection.py

---

Expected Output

• The application will launch in your default web browser.
• Turn on the camera permission in the browser the system will classify their freshness, display results, and store data in the MySQL database.

---

Application 2: QC Process Automation

---

Note on Running SpaCy and PaddleOCR Applications

This repository includes three independent applications—SpaCy, PaddleOCR, and QCprocessAutomation—designed to run in separate environments simultaneously. Follow the steps below to set up and run them correctly.

---

Clone the Repository

git clone https://github.com/Gopinathv19/FlipKartGrid6.0.git

---

Directory Structure

1. SpaCy Application

   • Located in the SpacyServer folder.
   • Handles NLP tasks such as MRP and DATE recognition and processing.

2. PaddleOCR Application

   • Located in the PaddleServer folder.
   • Facilitates OCR tasks for extracting text from images.

3. QC Process Automation Application

   • Located in the QCprocessAutomation folder.
   • Automates quality control tasks like product brand matching and data analysis.

---

Environment Setup

Each application has its own dependencies, and it's essential to create separate virtual environments for them.

Step 1: Set Up the SpaCy Application

1. Navigate to the SpacyServer folder:

   cd SpacyServer

2. Create a virtual environment and install requirements:

   python -m venv spacy_env
   source spacy_env/bin/activate    # On Linux/MacOS
   spacy_env\Scripts\activate       # On Windows
   
   pip install spacy
   
   python -m spacy download en_core_web_sm

3. Start the SpaCy server by running:

   python spacyServerTestApp.py

Step 2: Set Up the PaddleOCR Application

1. Navigate to the PaddleServer folder:

   cd PaddleServer

2. Create a virtual environment and install requirements:

   python -m venv paddle_env
   source paddle_env/bin/activate    # On Linux/MacOS
   paddle_env\Scripts\activate       # On Windows
   
   git clone https://github.com/PaddlePaddle/PaddleOCR.git
   cd PaddleOCR
   pip install requirements.txt
   
   cd ..
   
   pip install paddlepaddle
   

3. Start the PaddleOCR server by running:

   python paddleTextExtraction.py

---

Step 3: Set Up the QC Process Automation Application

1. Navigate to the QCprocessAutomation folder:

   cd QCprocessAutomation

2. Create a virtual environment and install requirements:

   python -m venv qc_env
   source qc_env/bin/activate    # On Linux/MacOS
   qc_env\Scripts\activate       # On Windows
   
   pip install -r requirements.txt

3. Run the QC Process Automation application:

   Execute the note book cell by cell 

---

Usage Instructions

1. Ensure that all three virtual environments are set up and activated in their respective directories.
2. Start the servers/applications individually:
   • SpaCy Application: Run spacyServerTestApp.py.
   • PaddleOCR Application: Run paddleTextExtraction.py.
   • QC Process Automation Application: Run QCprocessAutomation.py.
3. Ensure proper configuration of server ports to avoid conflicts.

---

Key Notes

• Each application must run in its own environment to avoid dependency conflicts.
• Verify that the dependencies specified in the requirements.txt file for each folder are installed correctly.

This modular setup ensures smooth execution of all applications, working seamlessly together to achieve the system's objectives. If you encounter any issues, refer to the respective folders for troubleshooting guides.

---

Key Technologies

YOLOv8: Utilized for detecting objects (products) in images, crucial for locating products and regions of interest on packaging.

EfficientNetB0: Used for classifying the freshness of detected products.

PaddleOCR: Facilitates the extraction of text from images, acting as a standalone OCR server.

SpaCy: Used for text recognition, analysis, and processing, particularly in understanding brand names, product details, and context from extracted text.

Sentence Transformer: Applied for matching and verifying brand names from text data,matching by the cossine similarity ,ensuring accurate identification during the QC process.

Flask: Used to set up a lightweight server for running the OCR model and serving text extraction requests.