Vehicle counting for traffic management using MATLAB

Published on 05 Jul 2021. Written by Sneha Kashyap

As we tend to all understand India could be a developing country. Increase in automobile sector comes with the event of development paralleled. This has led to widening in personal vehicles which cause a rise in congestion in large cities. So, we have a great need for a suitable traffic management system. This MATLAB project is all about to establish a program to make a traffic management system which is ready to detect live traffic flows are available in present traffic scenario in a lane. The system is to use already installed cameras in road networks with none additional controlled devices. It uses a strong division algorithm that discovers foreground pixels to cherish surrounding transport.

Read more..

SLNOTE


Project Description

Many types of traffic management systems are introduced to search out the distribution of vehicles on the road. The specific objective of this project is to see the amount and distribution of incoming traffic by using machine learning techniques. In this framework, a camera is installed and are used to capture the video of the highway or Lane. The video is recorded continuously in many frames, and every frame is compared to the primary clicked image. The overall number of cars present within the video is discovered using image processing algorithms. If the overall number of cars exceeds a predefined density, heavy traffic status is displayed as a message on the room.

This system is implemented in MATLAB with various modules. Four main modules are:

• Image acquisition
• RGB to grayscale Conversion
• Image enhancement
• Morphological operations

SLLATEST
Modules used in this project

The modules of the proposed system are:

Image acquisition:

• The two-dimensional function f(x, y)(here x and y are plane coordinates) is mostly associated with any picture. The amplitude of the image at any purpose say f is called intensity of the image.
• Then, these x and y values are get converted to finite discrete values to make a digital
• The intensity values are proportional to the radiated energy by a physical supply. Hence, pixel values should be nonzero and finite. e., 0< f(x, y) < ∞.

R to G Conversion:

All the colour images belong to RGB format. Grayscale images are always described by 8 bits. The pixel values are always represented using 256 levels varying from 0 to 255. One of the best methods of transformation is to take the intermediate of the contribution from all channel (R+B+C)/3. Nonetheless, since the perceived brightness is usually dominated by the Green part.

Image Enhancement

Image enhancement is the method acting for correcting digital images, therefore, all the results are more appropriate for display or advance analysis. For instance, we will eliminate noise, which is able to make it easier to spot the key characteristics. Image enhancement methods in MATLAB are used to obtain the grayscale version of the captured image with proper distinction and better quality.

Morphological operations

It contains operations on an image like:

Thresholding: Image thresholding could be an easy and effective method to differentiate a picture into foreground and background. It’s a segmentation process to isolate objects from the background.

Foreground Detection: The aim of foreground detection is to find changes occurring within the image frames. Foreground detection is finished to separate these changes going down within the foreground from the background. Here the foreground is detected using the mechanically generated threshold value taken using Otsu’s principle.

Vehicle Counting: There are several strategies presently in accustomed notice vehicles on the road like motion detectors, installation of lasers on both sides of the path, etc. Using MATLAB, the precise number of vehicles, are determined and therefore, the count is displayed employing a seven-segment display.

Software Requirements:

• MATLAB version R2016
• Graphical User Interface

Hardware Components:

• USB based web camera
• Hard Disc – 1 TB
• Memory – 8 GB RAM

Project Implementation

The algorithm behind this Project belong to the subsequent steps:

• Initiative is to begin the program.
• Get reference image, and therefore the image to be matched is continuously captured employing a camera that's installed at the junction.
• Capture image with vehicles
• The captured footage is converted from RGB to G
• The threshold value is found by Otsu’s principle.
• Find the difference between frames using a threshold.
• Apply Weiner filter to that to filter the blobs
• Convert to a binary image.
• Fill holes to the blobs
• Open all blobs having a locality greater than 200
• Determine the number of cars.
• Display the output image
• The count of vehicles is found and displayed.

Advantages:

• This method avoids the matter of traffic congestion by giving feedback on the vehicle count to control room.
• It's also a far better way of detecting the presence of transport on the road since it makes use of image data.
• The accuracy in calculation of waiting time because of the single moving camera.

 Future Scope:

• The hardware implementation would alter a project to use in real-time sensible conditions.
• A system is proposed to spot the vehicles as they go, giving preference to emergency vehicles and assisting in surveillance on an oversized scale.
• The climate don't seem to be taken into consideration which can affect the picture quality when it becomes blur or in heavy rains.

SLDYK
Kit required to develop Vehicle counting for traffic management using MATLAB:

• No kit required

Technologies you will learn by working on Vehicle counting for traffic management using MATLAB:

• MATLAB
• Image processing

Any Questions?