Evolution in Electrical and Electronic Engineering Vol. 2 No.

2 (2021) 703-708
© Universiti Tun Hussein Onn Malaysia Publisher’s Office

EEEE
Homepage: http://publisher.uthm.edu.my/periodicals/index.php/eeee
e-ISSN : 2756-8458

Vehicle Speed Estimation System by using

Matlab
Che Mohd Safwan Najmi Che Manan1, Wan Mahani Hafizah
Wan Mahmud1,2*
1
Department of Electrical Engineering, Faculty of Electrical and Electronics
Engineering,
Universiti Tun Hussein Onn Malaysia, Batu Pahat, 86400, Johor, MALAYSIA
2
Advanced Medical Imaging,
Universiti Tun Hussein Onn, Batu Pahat, 86400, Johor, MALAYSIA

*Corresponding Author Designation

DOI: https://doi.org/10.30880/eeee.2021.02.02.084
Received 04 July 2021; Accepted 01 August 2021; Available online 30 October 2021

Abstract: One of the biggest issues inside the UTHM community was the vehicle
over speeding issue. As a result of this problem, all car drivers within inside the
UTHM area exceed the speed limit. Besides that, The Vehicle Speed Estimation
System by MATLAB simulation software's was created to address this issue. This
system was built with the ESP32 Camera Module, and the video pictures were
processed with the MATLAB software. Any vehicle exceeding the speed restriction
within the UTHM area will be recorded and captured by this camera recorder system.
The MATLAB will display the offender vehicle's exact speed as proof that it is
exceeding the speed restriction in that location. This project will aid in the reduction
of excessive speeding inside the Universiti Tun Hussein Onn Malaysia area. This
project is also being created with the goal of lowering the accident rate in the UTHM
area and generating a more controlled and effective speed movement. As a result, the
entire UTHM community will take this matter more seriously and drive within the
speed limit within the UTHM area, as well as build a new better driving environment
within the UTHM area by paying closer attention to the car speedometer.

Keywords: Vehicle Speed, Speed Estimation System, MATLAB

1. Introduction
The Vehicle Speed Estimation technology was one of the most popular system that can help many
things in many situations. This technology has been introduced since past 4 years but all of these
technologies and system have been upgraded to become better and effective for the user. The current
speed detection was handheld by the police department that allow them to check the car speed and been
control by the police security department for any evidence on the scene. As the population grows, so
does the number of vehicles on the road. As a result, traffic surveillance systems based on video capture
*Corresponding author: wanmahani@uthm.edu.my
2021 UTHM Publisher. All rights reserved.
publisher.uthm.edu.my/periodicals/index.php/eeee
Che Mohd Safwan Najmi Che Manan et al., Evolution in Electrical and Electronic Engineering Vol. 2 No. 2 (2021) p. 703-708

and processing to evaluate speed have become critical concerns in preventing accidents. In Malaysia,
road accidents are one of the leading causes of death and injury. Aside from that, road accidents are
becoming more common, with a high percentage of individuals dying as a result [1]-[2].
According to the World Health Organization, around 1 million people are killed in traffic accidents
each year (WHO). This is a severe issue that needs to be addressed in future planning [3]-[4]. This
technology will detect over speeding vehicles on UTHM speeding area. The ESP32 Camera module
will monitor and record all the video footage on the speeding area. The MATLAB software will do the
speed calculation system. By doing this project, it will have many benefits to the driver to become more
alert while driving, thus will lead to decrement road accident percentage [5].
By the end of this project, we should have a Vehicle Speed Estimation system that can calculate
and estimate the vehicle's speed by using MATLAB software and Video Image Processing technique.
It is hoped that from this project, all drivers especially in UTHM may become more alert with the speed
limit, thus reducing the over speed cases which may leads to road accidents.
2. Materials and Methods
This study was carried out using MATLAB®. MATLAB is a programming environment for
engineers and scientists developed by Math Works. Next, the ESP32 Cam Module and his device can
monitor all the vehicle that cross into it in the video streaming session. Based on that video streaming
footage, the offender video was processed on the MATLAB as a video image processing and the
MATLAB will produce the actual speed of the offender vehicle.
This section goes through how to create a vehicle speed estimation algorithm in depth. The
algorithm for video image processing was developed using MATLAB. The created algorithm is tested
and simulated using a ready input video. Figure 1 shows how to create a vehicle speed estimation
algorithm in depth. The algorithm for video image processing was developed using MATLAB. The
created algorithm is tested and simulated using a ready input video. In order to compare the output,
different videos are used. For distinct input photos, however, separate zones of interest must be defined.

RGB to
Video Acquisition Video Frame Rate
Grayscale

Displacement and
Final Video Video Optical
Speed unit
Output Flow
Showing

Figure 1: Block Diagram of Algorithm Development

2.1 Video Acquisition

The Video Image Processing in MATLAB uses existing or imported video as input. The video can
be read using the MATLAB command videoFReader, and the input parameters will be shown and
captured in the MATLAB workspace. The photos can be seen in MATLAB using the videoFReader
command.

704
 Che Mohd Safwan Najmi Che Manan et al., Evolution in Electrical and Electronic Engineering Vol. 2 No. 2 (2021) p. 703-708

videoFReader = vision.VideoFileReader('Psm1.mp4');
v = VideoReader('Psm1.mp4');
%read video
videolink = 'C:\Program Files\MATLAB\R2016a\Psm1_Video';
paravideo = VideoReader(videolink);
t=paravideo.FrameRate;
T=paravideo.NumberOfFrame;
2.2 Video Frame Rate
%n=paravideo.FrameOfNumber;
It will display the frame rate of any video in the MATLAB can be read by using the video frame
OrVideoReader = vision.VideoFileReader(videolink);
rate command in the MATLAB. It will display the actual frame rate of the video. The ESP32 will
monitoring for the recording and capture the video footage on the scene. The ESP32 also will conduct
a live streaming video on the scene.
while hasFrame(v)
video = readFrame(v);
end
whos video

end

2.3 RGB to Grayscale

The four types of images are binary, grayscale, RGB, and indexed images. RGB images are color
or images in which each pixel has a red, green, and blue color combination. The pixel value in a
grayscale image varies from 0 to 255, with shades of grey in between. The photographs must be
converted from RGB to grayscale before processing. The rgb2gray function is used as a result. Using
this function, 24-bit RGB values are converted to 8-bit Grayscale values.

%OrVideoReader = vision.VideoFileReader( videolink, ....

%'ImageColorspace', 'RGB',...
%'VideoOutputDataType', 'unit8');
2.4 Video Optical Flow
It is a pattern that depicts how objects, surfaces, and edges appear to move. The observer's relative
motion to the scene can also help to construct the imagined scene. It is also possible to establish a
distribution of apparent velocities of any movement and a brightness pattern in a picture.

%create optical flow

opticalFlow.Method = 'Lucas-Kanade';
converter = vision.ImageDataTypeConverter;
OpticalFlow = vision.OpticalFlow('ReferenceFrameDelay', 3);
OpticalFlow.OutputValue = 'Horizontal and vertical components in complex form';
shapeInserter = vision.ShapeInserter('Shape', 'lines', 'BorderColor', 'Custom',
'CustomBorderColor', 255);

705
Che Mohd Safwan Najmi Che Manan et al., Evolution in Electrical and Electronic Engineering Vol. 2 No. 2 (2021) p. 703-708

2.5 Displacement and Speed unit showing

The speed (velocity) is the rate of change of distance (displacement) with time. The speed in this
system will showing in km/h.
disp(cen); %show...
disp(size(cen));%show....
speed_str = [num2str(a),'Km/h'];

2.6 Final Output Video

The video image processing will show the final output in MATLAB. It shows the vehicle tracking
and speed showing on that video processing in km/h.

%release video reader and writer

release(OrVideo);
release(OfVideo);
release(BwVideo1);
release(BwVideo);
release(ReVideo);
release(OrVideoReader);
3. Results and Discussion
Figure 2 shows the original video that been import to the MATLAB as the input video for the Video
Image Processing. The video depicts the car's movement and motion within the UTHM speeding zone.
The ESP32 Cam Module will record all video footage on the scene, which will then be processed in the
MATLAB software to determine the car's true speed.

Figure 2: The original image that export from the ESP32 live streaming video footage

Figure 3 shows the displaying the calculated speed from the MATLAB software. The MATLAB
Video Image Processing will show the actual speed of the car in Km/h. The real speed of the car was
46.7 km/h, according to this result. The live streaming video of the ESP32 Cam Module was used to
record the entire video. The ESP32 Cam Module's live streaming video may also record and capture
every scene that passes across it.

706
 Che Mohd Safwan Najmi Che Manan et al., Evolution in Electrical and Electronic Engineering Vol. 2 No. 2 (2021) p. 703-708

Figure 3: Displaying the calculate speed from the MATLAB software

Figures 2 and 3 show the original image export as well as the MATLAB and the final speed output
from the MATLAB. The ESP32 Cam Module will record all video footage on the scene, which will
then be analysed in MATLAB software to determine the car's true speed. The exact speed of the car in
Kilometres per hour will be displayed using MATLAB Video Image Processing. The real speed of the
car was 46.7 km/h, according to this result. The ESP32 Cam Module's live streaming video may also
record and capture every scene that passes across it.
The speed data comparison between MATLAB and the real speed from the car speedometer is
shown in Table 1. The speed range for the first test was 10-20 km/h, and the MATLAB speed was 18.28
km/h, whereas the real speed from the real car speedometer was 20.00 km/h. Then, at a speed of 20-30
km/h, the MATLAB speed was 28.11 km/h, but the true speed from the car speedometer was 30.00
km/h. Finally, the speed between 30 and 40 km/h was tested, and the MATLAB result was 38.80 km/h,
whereas the real-time speedometer on the car indicated 40.00 km/h. Although there were a few technical
challenges with the device, all of the results from the comparison of MATLAB speed output and real
car speedometer speed were close to success.
Table 1: The speed data comparison between MATLAB and Real speed from the car speedometer

No Speed Test MATLAB Final Speed Real speed from car

(Km/h) Result (Km/h) speedometer (Km/h)

1 10-20 18.28 20.00

2 20-30 28.11 30.00
3 30-40 38.80 40.00

4. Conclusion
In this study, the traffic conditions and scene on the road in the UTHM community area were used.
Any vehicle's speed can be tracked using the developed algorithm. The proposed algorithm
accomplishes the objectives. The system has a flaw in that it can recognize but cannot accurately track
their speed. When an image or video is captured, a number of factors might influence the quality and
noise of the image or video. One of them is when there is bad weather, such as rain, which causes the
picture and video recording results to blur and degrade in quality. Besides that, one of the factors that
affected the video or image recording result was the darkness. This issue will cause the outcome to be
overly dark, and the vehicle's image will not be as clear as it would be on a normal sunny day. Using
these techniques, vehicle detection can be improved and errors reduced.

707
Che Mohd Safwan Najmi Che Manan et al., Evolution in Electrical and Electronic Engineering Vol. 2 No. 2 (2021) p. 703-708

Acknowledgement
The authors would like to thank the Faculty of Electrical and Electronic Engineering, Universiti
Tun Hussein Onn Malaysia for its support.
References

[1] M. Rath, “Smart Traffic Management System for Traffic Control using Automated
Mechanical and Electronic Devices,” IOP Conf. Ser. Mater. Sci. Eng., vol. 377, no. 1,
2018, doi: 10.1088/1757-899X/377/1/012201.
[2] Gupte, S.; Masaud, O.; Martin, F,K,R; Papanikolopuolos, N,P. Detection and
classification of vehicles, IEEE Trans, Intell, Transp, Syst, 2002, 3, p. 37-47.
[3] Moazzam, Md.G., Haque, M.R. and Uddin, M.S. (2019) Image-Based Vehicle Speed
Estimation. Journal of Computer and Communications, 7, p. 1-5.
[4] Different Approaches for Motion Estimation”, E.F.( 4th-6th June 2009 ), International
Conference on control, automation, communication and energy conservation -2009.
[5] Lazaros Grammatikopoulos, George Karras, Elli Petsa, E .F.( November,
2005),“Automatic Estimation of Vehicle, Speed from Uncalibrated Video Sequences”
International Symposium on Modern Technologies, Education & Professional Practice
in Geodesy and related fields, Sofia, p.03 – 04

708