Skip to content

An implementation using Neural Network and Computer Vision concepts to analyse traffic density given video and photos

Notifications You must be signed in to change notification settings

Ajayraj-M-2003/Traffic-Density-Analyser

Repository files navigation

Traffic-Density-Analyser

In this project we focus on analysing the traffic density using computer vision concepts and cnn model. We work on taking a photo or video as input and give a feedback whether the road traffic is high or low. This operation was done using pretrained model , i.e. Yolov8 for vehicle detection.

Contributors:

Ajayraj M | Praveen N | Reshiha R G | T Vinay Kumar |

Datasets referred:

  1. https://www.kaggle.com/datasets/yusufberksardoan/traffic-detection-project
  2. https://www.kaggle.com/datasets/farzadnekouei/top-view-vehicle-detection-image-dataset

About

An implementation using Neural Network and Computer Vision concepts to analyse traffic density given video and photos

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published