Skip to content

gourab9817/Flipkart_grid_6.0

Repository files navigation

Flipkart_grid_6.0

Hexagone

Demo Video Link: https://youtu.be/8s3sAIvPJv0?feature=shared

Hexagone

Step-by-Step Guide to Setting Up and Running the Application

  1. Clone the Repository

    https://github.com/gourab9817/Flipkart_grid_6.0.git
  2. Create a Virtual Environment

    python -m venv Grid
  3. Activate the Virtual Environment

    • On Windows:
      Grid\Scripts\activate
    • On Linux/macOS:
      source Grid/bin/activate
  4. Install Required Packages

    pip install -r requirements.txt
  5. Run the Application

    python manage.py runserver

Freshness Detector

Hexagone

  • Predicts the freshness and identifies the class (type) of fruits and vegetables.
  • Utilizes a model I trained using a dataset from Kaggle with MobileNetV2 as the base model in TensorFlow.
  • Hexagone

Feature Extractor

Hexagone

  • Extracts product details such as MRP, EAN, manufacture date, and expiry date using OCR powered by Pytesseract.
  • Processes the text to validate the expiry date of the product. Hexagone

Object Detection

Hexagone

  • Counts and highlights products within an image.
  • Employs the EfficientDet model from TensorFlow Hub. Hexagone

Dataset For Freshness Detector

Download the dataset from Kaggle.

Fresh fruits- fresh banana, fresh apple, fresh orange, fresh mango and fresh strawberry.

Rotten fruits- rotten banana, rotten apple, rotten orange, rotten mango and rotten strawberry.

Fresh vegetables- fresh potato, fresh cucumber, fresh carrot, fresh tomato and fresh bell pepper.

Rotten vegetables- rotten potato, rotten cucumber, rotten carrot, rotten tomato and rotten bell pepper.

Pre-trained Model / Architecture for Object Detection

Tesseract OCR Engine for Optical Character Recognition