Professional Documents
Culture Documents
Abstract—In this paper, a vehicle safety detection and alarm remote alarm in the mobile phone application. If the alarm
system can be used to detect the inside environment of the information is not be released in time, or network
vehicle. When it is found that the inside environment of the communication has problems, the equipment will be set up
vehicle after parking is bad, and there are people in the car, it to contact the telephone (including 110), send alarm and
can alarm through this system, so as to ensure the safety of rescue messages containing location information and risk
personnel. factors.
Users can check the environmental information and
Keywords- Vehicle safety detection; Safety of personnel;
location information of the vehicle at any time through the
Alarm
mobile application terminal. When someone is stuck in the
car and danger occurs, the mobile phone will alarm by
I. INTRODUCTION notification. Web server is mainly used for information
Traffic safety is a topic of great concern to the whole interaction between hardware devices and mobile
society. The government has issued standards and policies, applications.
and the automotive industry continues to develop various The danger of children being trapped in cars occurs
safety detection devices, such as airbags, ABS anti-lock mainly during the day, and almost all of them die of heat
braking system, tire pressure detection, to improve driving stroke; adults die of exhaust gas poisoning. Therefore, this
safety. hardware device mainly aims at the above situation to
However, there is no specific safety precautions for those propose the solution.
people, especially children, who remain in the vehicle after This paper mainly discusses the hardware and software
the vehicle is parked. Children do not have some basic design of vehicle safety monitoring and alarm equipment, as
behavioral skills and can not alarm themselves when trapped well as related experience and experience.
in the car for help. So every year, especially in summer,
there are always many children forgotten and stranded in the III. HARDWARE STRUCTURE DESIGN OF EQUIPMENT
car, resulting in heatstroke and death tragedy. Not only The hardware structure of vehicle safety detection and
children but also adults may die of exhaust gas poisoning q p
alarm equipment g
is shown in figure 2.
due to air conditioning switching on during parking.
The system is designed and developed for the safety of
people who are stuck in the vehicle, especially for children.
II. OVERVIEW OF FUNCTION
The system is mainly composed of three parts, namely,
the car safety monitoring and alarm hardware equipment,
web server, and android mobile phone application, its
structure diagram
g as shown in figure
g 1.
10
and there is still danger, then through the GSM module Here it use the Perceptual Hashing Algorithm for image
contact 110 alarm. alignment. A corresponding fingerprint string is generated
When adults are trapped in the car, the danger usually for each frame image. The larger the difference between the
occurs in the parking state, due to the start of air two fingerprint images, the less close the image is. The
conditioning caused by the tragedy of exhaust gas poisoning, simplified fingerprint string generation method is as follows:
so it is necessary to analyze the gas in the car. Carbon First, the average gray value of 64 pixels is calculated, and
monoxide gas contains a large number of exhaust gas, so the then each pixel is compared with the average value. If it is
equipment uses a carbon monoxide detection sensor, when greater than the average value, it is marked 1 at the
the concentration of the gas exceeds the standard, as a corresponding position of the fingerprint string, otherwise it
dangerous situation, and will produce alarm information. is marked 0. Comparing the fingerprint coding of two
images, the number of the same data at the corresponding
B. Image Data Processing position in the two fingerprint strings is counted, which is
The device uses OV7670 image sensor to acquire image the similarity of the two images.
data in a car at regular intervals. If there are people staying
in the car, because of the activities of the people, the image V. CONCLUSIONS
data will be quite different. By comparing the difference of The vehicle safety detection and alarm system designed
the image data, we can judge whether there are people in this paper can effectively avoid people staying in the car
staying or not. when parking, because of heat stroke or exhaust gas
The image data processing flow of this system is shown poisoning and unfortunate situation. In this paper, the
in Figure
g 4. architecture and design of hardware devices are described
and introduced in detail.
However, the system can not be applied to detection of
personnel in dim light (such as night). In addition, video
compression is very serious, and there are some errors in
image matching. Therefore, the hardware and software of the
system have room for improvement.
REFERENCES
[1] Zhou Wenhai, Li Guanghui, Yang Lin, “Alarm Deive Control
System Design based Children Stranded in Auto”, Journal of
Liuzhou vocational & Technical College, 2015, 33: 72-74. (in
Chinese)
[2] You Dazhang, Zhang Jiangang, Gan Yong, “Bitmap Image
Grayscale Method and Programming Implementation”,
Journal of Guangxi University of Technology, 2004, 15:23-
25. (in Chinese)
[3] Li Huimin, Fan Jiming, Yang Xiao, “Design of image
acquisition and display system based on STM32 and
Figure 4. Image processing flow OV7670”, Transducer and Microsystem Technology, 2016, 9:
114-117. (in Chinese)
When setting the OV7670 image sensor, the QVGA
format is adopted, that is, the resolution of the output image
is 320 * 240, but only 8 * 8 pixel color values are stored
evenly when collecting the image information, and the color
values are in RGB 565 color mode.
Generally, the relationship between image similarity and
color is not very large, so it needs to be processed as gray
scale image to reduce the complexity of later calculation.
Gray scale processing is needed to consider gray scale. The
gray scale represents the level of brightness between pixels
from the darkest to the brightest. Let color be a pixel color
value in RGB565 format.
First, get the RGB component value according to the
formula below.
gm_red=(color&0xF800)≫8
gm_green=(color&0x07E0)≫3
gm_blue=(color&0x001F)≪3
Then get the 32 gray scale value by the following
formula.
gray=(gm_red*77+gm_green*150+gm_blue*29)≫11
11