Map Matching Enhancement by Deep Learning

This project provides a web interface for uploading vehicle trajectory data, processing it with a Hidden Markov Model (HMM), and displaying the results on an interactive map.

Prerequisites

• Python 3.7+
• pip (Python package installer)
• git

Setup and Running

Follow these steps to set up and run the project:

1. Clone the Repository:

   https://github.com/shivas1516/map-matching.git
   cd map-matching-hmm

2. Create a Virtual Environment:

   python -m venv venv

3. Activate the Virtual Environment:

   • On Windows:

     venv\Scripts\activate

   • On macOS and Linux:

     source venv/bin/activate

4. Install the Required Packages:

   pip install -r requirements.txt

5. Run the Flask Application:

   python app.py

6. Open the Web Application:

   Open your web browser and go to http://127.0.0.1:5000/.

Usage

1. Prepare Your CSV Data File:

   Ensure your CSV file contains at least lat (latitude) and lon (longitude) columns, along with optional additional columns like timestamp.

2. Upload and Process Data:

   • On the web interface, click on the "Choose File" button and select your CSV file.
   • Click the "Run Map Matching" button to upload the file and process the data.
   • The processed map will display both the original and matched points.

Example

Here's an example of how your CSV file might look:

Download as CSV

https://docs.google.com/spreadsheets/d/1rFiZ5-uP_227DsbvBewnovyaNmUj9N3ahw3tFGkjjFI/edit?usp=drivesdk

Save this as trajectory_data.csv and upload it through the web interface to see the results.