Proceedings of 2012 4th International Conference on Machine Learning and Computing
IPCSIT vol. 25 (2012) © (2012) IACSIT Press, Singapore
Design of a Smart Car based on Wireless Image Acquisition
Technology
FAN Yuezhen +, ZHENG Fu, DING Feng and QIAN Hua
School of Technology, Beijing Forestry University, Beijing 100083, China
Abstract. Design of a smart car based on wireless image acquisition technology is introduced. The smart
car is controlled by MCU MC9S12DG128. It receives control commands to carry a camera to different sites
to capture images. There is a CMOS camera and a ZigBee RF module installed on the smart car. The CMOS
camera can obtain static images. Then the acquired image data are stored in MC9S12DG128. And the image
data are transferred to a remote monitor on PC through wireless ZigBee RF module based on IEEE
802.15.4/ZigBee protocol. Finally, the smart car can be used to capture and send images of special
environments, such as toxic environment and narrow spaces, to a remote monitor center.
Keywords: Smart car, CMOS camera, ZigBee, Image acquisition
1. Introduction
In recent years, lots of disasters occurred on the earth. But in the rescuing sites, rescuers can’t get access
to dangerous environments because of radiation, poisonous gas, or narrow spaces. The Fukushima Daiichi
nuclear disaster led to global panic, which is nuclear meltdowns and releases of radioactive materials at the
Fukushima I Nuclear Power Plant, following the Tohoku earthquake and tsunami on 11 March 2011. At that
time, rescuers could not get close to the disaster site in fear of the nuclear radiation. To solve the problem, a
smart car based on wireless image acquisition technology was constructed. It can be used to acquire and send
images of special environments though wireless ZigBee network to a remote monitor center.
2. System Analysis
The wireless image acquisition system is composed of several parts. The main controller is based on
Freescale MC9S12DG128 [1], which is a 16-bit micro controller and has enough peripherals, such as PWM,
UART and so on. A keyboard is connected as the I/O device. A PWM module output to the servo to control
the direction of front-wheel. Another PWM module is outputted to the MC33886 driver to drive the DC
motor at the rear-wheel. The photoelectric encoder installed in the rear-wheel is used to measure the velocity
for a feedback control [2, 3]. There is a CMOS camera OV7620 controlled by OV529, capturing pictures.
The wireless ZigBee module CC2530, connected to the UART interface of the MCU, is used to send image
data to the remote monitor and receive commands from it. The interconnection among these modules is
shown in Figure 1.
+
Corresponding author. Tel.: +86-010-62336221.
E-mail address: fanyuezhen@163.com.
194
Fig. 1: The structure of the wireless image acquisition system
3. Hardware Design
In order to complete the system’s function, hardware circuits for CMOS camera [4], ZigBee RF module
and smart car was designed.
3.1. CMOS Camera
The image sensor module for our platform is OmniVision OV7620 combined with embedded DSP
OV529. The OV7620 image sensor is a low voltage CMOS sensor that provides the full functionality of a
single-chip VGA camera and image processor, as shown in Figure 2. The OV7620 provides full-frame, subsampled or windowed 8-bit images in a wide range of formats, operating at up to 30 frames per second (fps),
controlled through the Serial Camera Control Bus (SCCB) interface [5]. The OV529 Serial Bridge contains
an Embedded JPEC CODEC and controller chip that can compress and transfer image data from the Camera
Sensor to an external device [6]. The OV529 performs all imaging function like white balance, downsizing
and compressed image to JPEG format. And the image sensor module connects with the processor through
UART.
U9
D[0..15]
R17 10K
AGCEN3
+5V
SVDD 1
AVDD 8
14
44
DVDD 29
32
+3.3V
48
6
7
15
43
30
31
C62 0.1uF
5
C63 0.1uF
10
C64 0.1uF
11
13
RESET 2
FREX 4
SBB 12
R18
10K
9
47
27
R19
10K
AGCEN/RAMINT
SVDD
AVDD
AVDD
AVDD
DVDD
DOVDD
SGND
AGND
AGND
AGND
AGND
DGND
DOGND
VREQ
VRS
VCCHG
VTO
RESET
FREX
SBB
PWDN
MID
XCLK1
R20
OV7620
Y2
C65
1
Y0/CBAR
Y1/PROG
Y2/G2X
Y3/RAW
Y4/CS1
Y5/SHARP
Y6/CS2
Y7/CS0
UV0/GAMDIS
UV1/CC656
UV2/QVGA
UV3/ECLKO
UV4/SLAEN
UV5/MIR
UV6/BPCLR
UV7/B8
PCLK/OUTX2
VSYNC/CSYS
FODD/SRAM
CHSYNC/BW
HREF/VSRAM
SIO-1
SIO-0
XCLK2
41
40
39
38
37
36
35
34
26
25
24
23
22
21
20
19
33
16
17
42
18
45
46
28
D0
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
D11
D12
D13
D14
D15
PCLK
VSYNC
SRAM
CHSYNC
HREF
SCL
SDA
1M
2
6.14MHz
33p
C66
33p
Fig. 2: The schematic of the CMOS camera
3.2. ZigBee Module
A ZigBee transceiver, CC2530, was used as a RF module to communicate with the PC. Thus, the car
model can transmit image data to the PC. And the PC also can send control commands to the car model. The
195
CC2530 is a true system-on-chip (SoC) solution for IEEE 802.15.4, Zigbee and RF4CE applications. It
enables robust network nodes to be built with low costs. The schematic of CC2530 is shown in Figure 3.
1
744
DCOUPL
RBIAS
GND
728
2
32
2 L261
2
1
Ind
2N0
1
22
23
40 639
30 747
41
604
21
1 2 1 2
2
P4
R301
56K
C401
1U
3
276
X1
736
1
4
32M
738
X2
1
32768
C221
27p
1 2 1 2
C231
27p
1 2 1 2
1 2 1 2
1 2 1 2
1
Header 2
C262
1p
33
CC2530
2
1
C253
000
C331
15p
C321
15p
P2
VDD
Reset
P1.2
P2.0
1
3
5
7
9
11
13
15
17
19
2
4
6
8
10
12
14
16
18
20
P0.5
Header 10X2
Fig. 3: The schematic of CC2530
3.3. Motor Driving Module
The smart car is driven by a DC motor in the rear-wheel; its direction is controlled by a servo in the
front-wheel [7].
To obtain a stable speed of the motor, feedback mechanism PID was used to eliminate environment
interference such as non-stable battery voltage and uneven road friction. So an integrated photoelectric
encoder was used to get the speed of the motor. The signal is connected to PT7 of the MCU.
RS-380SH motor was used. It works at voltage 7.2V and no-load current 0.5A, whose specified rotating
speed is 16200 r/min. MC33886 is a monolithic H-Bridge ideal for fractional horsepower DC-motor and bidirectional thrust solenoid control. Two chips of MC33886 are paralleled to drive the motor, which can
provide greater driving force. And the MC33886s are controlled by a PWM from the MCU. Changing the
duty cycle of the PWM, then the speed of the motor was adjusted [8]. The schematic of the driving circuit is
shown in Figure 4.
196
SMA_SMD
5
4
3
2
XSOC32M_Q2
734
2 1 2 1
L251
Ind
000
21
2
1
21
2
C254
2p2
2 1 2 1
XSOC32M_Q1
1 2 1 2
18p
C261
1p
L252
Ind
2N0
2 1 2 1
P2_3
1 2 1 2
26 724 1
21
18p
P2_4
1 2 1 2
25 725 1
C251
12
7321
1
1
RF_N
2
C252
1
RF_P
731
2
P0.4
P0.5
P0.6
P0.7
Reset
P2_0
P2_1
P2_2
P1_0
P1_1
P1_2
P1_3
P1_4
P1_5
P1_6
P1_7
P0_0
P0_1
P0_2
P0_3
P0_4
P0_5
P0_6
P0_7
RESET_N
C1
2U2
VDD
2
C311
100n
P3
21
24
27
28
29
31
1
Header 10X2
36
35
34
11
9
8
7
6
5
38
37
19
18
17
16
15
14
13
12
20
C272
220p
1
P2.0
P2.1
P2.2
P1.0
P1.1
P1.2
P1.3
P1.4
P1.5
P1.6
P1.7
P0.0
P0.1
P0.2
P0.3
P2.1
P2.2
P1.4
P1.5
P1.6
P1.7
C271
100n
2
P1.3
P1.0
AVDD5/AVDD_SOC
AVDD3
AVDD2
AVDD1
AVDD4
AVDD_GUARD
2
2
4
6
8
10
12
14
16
18
20
DVDD
AVDD_DAEG
DGND_USB
USB_M
USB_P
DVDD_USB
C241
100n
2
1
3
5
7
9
11
13
15
17
19
U1
C211
100n
1 2 1 2
10
39
1
2
3
4
1 2 1 2
C101
100n
P1
P0.4
P0.1
P0.2
P0.3
P0.0
P1.1
P0.6
P0.7
1 2 1 2
1 2 1 2
1 2 1 2
C391
1U
L1
1
2
Bead
102
C255
000
Fig. 4: The schematic of the motor driving circuits
4. Software Design
To achieve the system, 3 types of specified software were developed. One on CC2530 was the ZigBee
protocol stack to communicate between the smart car and the remote controller; one was the smart car
control software based on MC9S12DG128; another one was the PC monitor developed with Microsoft MFC.
A set of protocol of commands were specified so that the PC monitor and the smart car must obey with it to
communicate with each other accurately.
4.1. CC2530 Software
In a ZigBee network, there are 3 types of nodes; they are coordinator, router, and end device [9]. Each
node can be configured as one of these above. So the node connected to the PC was configured as a
coordinator; the one installed on the car was configured as an end device.
When powered on, the coordinator set up a ZigBee network, and then the end device enters the
established network. Since then, the PC monitor sent commands to the smart car to take pictures or move
around; the smart car controlled the camera to take a picture and sent the image data to the PC monitor. As is
specified [9], packet size of ZigBee is 127 Bytes; despite for necessary controlling bytes, the payload of each
packet is less than 100 bytes. As a result, we separated the image data into 64 bytes per packet to transfer.
The flowchart of the wireless communication procedure is shown in Figure 5.
Start
Abandon cmmand
N
Coordinator setting up
ZigBee network
Move command
Take picture command
End device entering
network
Set servo angle
Send COMS camera
command
Set motor speed
Store the data in memory
Communication
OK?
Y
Calculate the packet
number
Smart car waiting for
command from monitor
Send the image data
Smart car receiving
command
Fig. 5: The flowchart of the wireless communication procedure
197
4.2. Smart Car Control
The car receives the commands through the wireless ZigBee CC2530 module. According to different
commands, the car moves can move back and forth, turns around, takes pictures and sends image data.
First of all, the length of the servo rod was adjusted to an optimal value. Thus, the servo has the most
extent of sensibility. The servo is controlled by PWM. And the servo's rotation is proportional to the pulse
width. However, the relationship between the turning radius of the car models and steering angle of the servo
is nonlinear [10]. The relationship between them was obtained through lots of experiments, and then the data
was plotted in MATLAB, as shown in Figure 6. They were stored in the memory as a table. When the car is
set to a specified turning radius, the MCU will search the table for the corresponding PWM duty cycle and
outputted it.
Traditional PID control strategy was adopted to control the speed of the motor [11]. Compared to openloop control, the PID feedback control can keep the motor maintaining a stable speed. The PID parameters
were tuned through lots of experiments.
Fig. 6: The relationship between turning radius and servo angle
4.3. PC Monitor Center
A monitor program was developed with the Microsoft MFC, as shown in Figure 7(Image Captured in our
Lab.). The upper right combo list and button were used to control the COM port; the middle two buttons
were used to take and show pictures, acquired by the smart car; the bottom right 4 buttons were used to
control the movement of the car. Then the bottom box showed the state of the program dynamically. Because
the image is JPEG format, there is a decoder module in the program.
Fig. 7: The MFC monitor program
198
5. Conclusions
By designing the CMOS camera, wireless ZigBee module and smart car control module, a wireless
image acquisition system was developed. In experiments, as the car moved around, it can acquire and send
image of its surroundings to the PC monitor to display.
6. Acknowledgements
This paper was supported by Chinese Universities Scientific Fund (BJFU-YX2010-8).
7. References
[1] Zhuo Qing, Huang Kaisheng, and Shao Beibei, “Learn to make smart cars.” Press of Beijing University of
Aeronautics and Astronautics,2007.
[2] G. Yue-hua, "Design of Autonomous Tracing System in Intelligent Vehicle Based on Infrared Photoelectric
Sensor," Semiconductor Optoelectronics, vol. 1, 2009.
[3] X. W. Hao and H. F. Sun, "Tracking Control Algorithm Based on Photoelectric Sensors for Smart Car System,"
Iccsse 2009: Proceedings of 2009 4th International Conference on Computer Science & Education, pp. 587-591,
2009.
[4] T. Schweers, "Smart hyper - a hybrid concept-car with high innovative potential," Vehicle Concepts for the 2nd
Century of Automotive Technology, vol. 1653, pp. 243-254, 2001.
[5] Z. H. Wang and Y. Y. Ye, "Image Recognition Technology and Optimized Control Algorithm of the Self-tracing
Smart Car Based on CMOS Camera Sensor," Icems 2008: Proceedings of the 11th International Conference on
Electrical Machines and Systems, Vols 1- 8, pp. 2096-2101, 2008.
[6] X. Q. Wang, et al., "Route Identification and Direction Control of Smart Car Based on CMOS Image Sensor,"
2008 Isecs International Colloquium on Computing, Communication, Control, and Management, Vol 2,
Proceedings, pp. 176-179, 2008.
[7] P. K. Chang, et al., "Integrated smart car navigation and voice control system design," Iecon 2006 - 32nd Annual
Conference on Ieee Industrial Electronics, Vols 1-11, pp. 23-28, 2006.
[8] M. F. Wang and Z. J. Jiang, "The Design and Comparison of DC Motor Drive Circuit in Smart Car Competition,"
2009 Ieee International Conference on Intelligent Computing and Intelligent Systems, Proceedings, Vol 2, pp.
325-328, 2009.
[9] ZigBee Alliance. ZigBee Specification.ZigBee Document 053474r07, Version1.1, September5th, 2005.
[10] S. H. Ma and W. Li, "Non-linear PID Control Method with Feed-forward in Smart Car," Ccdc 2009: 21st Chinese
Control and Decision Conference, Vols 1-6, Proceedings, pp. 462-465, 2009.
[11] Z. L. Lu, et al., "Intelligent Control Research based on the Smart Car," 2nd IEEE International Conference on
Advanced Computer Control (Icacc 2010), Vol. 2, pp. 292-297, 2010.
199