egemengulpinar / Car-Detection-Using-with-MATLAB

This repo made for detection cars, trucks and motorcycles using with highway traffic cameras video examples

Car Detection Using with MATLAB

This repo made for detection cars, trucks and motorcycles using with highway traffic cameras video examples. It created for small systems, it has not need deep learning algorithms, machine learning methods or large traffic cameras video datasets. I have run this project on my own computer. 1920x1080 resolution executed with good performance. Also, small resolution video files have executed more effective and fast.

Car Detection using With MATLAB has not reach to %100 accuracy. But it performs good performance on small systems and executed with quick debug and process time.

Note: I have prepared explanation video to understanding easly of my project https://drive.google.com/file/d/1oL9pCEocH4AmkcmUrowI9Hj5db62OXwk/view?usp=sharing

Note: I have prepared paper with my teammate, you can reach it with below the link https://drive.google.com/file/d/1qcnQ4bXVvOcKbETRIP9dAdW0x_BFyJR5/view?usp=sharing

Created for Whose?

• Created for who have small system
• Created for who wants to quick and fast process time on detection system.
• Created for who needs average accuracy and count number

Demo

Demo 2

Screenshots

Using & Run

Clone the project

  git clone https://github.com/egemengulpinar/Car-Detection-Computer-Vision.git

Open MATLAB for CarDetectionProject.m import all project files.

It can be executed using with videoplayback_short.mp4 or car2.mp4 files. (You can use another stable traffic camera video)

Video = VideoReader('videoplayback_short.mp4');

or

Video = VideoReader('car2.mp4');

Important Notes

This project must to optimize for input video. There is few notes below. This project who used, it should be change vision.ForegroundDetector and vision.BlobAnalysis parameters.

• In MATLAB function vision.ForegroundDetector detects foreground using Gaussian Mixture Models.

• Then, Morphological Operation strel('square',3) executed to Remove Noises.

• Locating the object with vision.BlobAnalysis