Traffic-Density-Estimation

Implementation of deep learning based real time traffic density estimation.

Overview

This project is a deep learning based approach to estimate the traffic density on a road in real time. Two images or video flows are feature matched and stitched together and then the total vehicles in that image/video are detected. Based on the number of vehicles, size of bounding boxes and the total occupancy of the road in that image/video frame, the traffic density is estimated for that frame.

Concepts Utilised

• ORB Feature Matching
• Image Stitching
• Homography matrix using RANSAC
• YOLOv3 object detection architecture
• Motion Detection using Image Substraction

Installation & Implementation

Step 1: Clone this repository.

git clone https://github.com/parthmalpathak/Traffic-Density-Estimation.git

Step 2: Run the relevant file to visualize results.

cd Traffic-Density-Estimation/Code
Traffic Density Estimation.ipynb

Note: User will have to pass 2 images which depict the left and right part of the image scene for successful functionality of this project

Resultant Output

Copyright

Author @Parth Malpathak