Exeley Inc.

Shebzukhov, Yuri

Yubian, Wang

International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
1 
Research and Realization on the key technology of  
Motion Image Matching Based on UAV 
Yuri Shebzukhov 
Department of the International Relations 
Belarusian State University of Transport, Republic of 
Belarus  
34, Kirova street, Gomel, 246653, Republic of Belarus 
e-mail: oms@bsut.by 
Wang Yubian*  
Department of Railway Transportation Control 
Belarusian State University of Transport 
34, Kirova street, Gomel, 246653, Republic of Belarus 
e-mail: alika_wang@mail.ru 
 
Abstract—With the development of communication network 
and computer software and hardware technology, especially 
the emergence of high-precision and high-resolution image 
sensors, the photography and measurement technology of 
aerial image have played an increasingly important role in 
today's geological survey. The traditional aeronautical 
measurement is carried out by a large manned aircraft. The 
collected measurement information has a large capacity and a 
wide shooting range, which is suitable for large-area 
operations. This type of measurement requires high hardware 
requirements and is expensive, and it is not suitable for small 
areas. UAV aerial measurement system is used to measure the 
small areas, UAV is in small size, with great flight fluctuation, 
but the data image collected is not accurate enough. In this 
paper, the common algorithms of image fast matching are 
compared to conduct in-depth research on gray-level matching 
and feature-based matching, and SIFT feature matching 
algorithm based on feature matching is proposed to obtain the 
measurement image as consistent as possible with the actual 
scene. The main features of the object to be measured can be 
obtained through the actual surface area image measurement 
test, which is of practical significance in the practical 
low-altitude small area surveying and mapping. 
Keywords-UAV; Moving Image; Feature Detection; Image 
Matching; Image Pre-Processing 
I. INTRODUCTION 
UAV is named quad-rotor aircraft or quad-rotor UAV [1], 
It is a four-propeller sky aircraft with cross-shaped propellers. 
The UAV can be used to take aerial photos or record video 
with an optical camera or miniature video recorder. UAV 
photography measurement system is established on the basis 
of unmanned aerial vehicle (UAV) mobile platform, it is a 
kind of advanced measurement technology to achieve high 
spatial resolution remote sensing image data, it is playing the 
very major role in the geological landform measurement, 
disaster reduction, disaster prevention, emergency rescue, 
emergency treatment, post-disaster reconstruction and so on. 
At present, in the aspect of UAV it mainly relies on the 
image aerial camera to collect the image inland and aboard. 
The traditional measurement camera is not only expensive, 
but also needs to carry out film image scanning to obtain the 
digital image. Its shooting quality is low and the 
measurement takes a long time. With the development of the 
UAV aerial survey technology, storage and transmission 
technology, using the measurement type camera CCD image 
acquisition has been widely used, the CCD camera has the 
advantages of a low price, sensors work stability, high 
sensitivity, the camera CCD cannot direct measurement, the 
difference of image distortion correction is bigger, so before 
shot aircraft must be matching the calibration. 
DOI: 10.21307/ijanmc-2019-014
International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
2 
II. PREPROCESSING OF UAV IMAGE  
As the UAV photography system is equipped with a 
non-professional measuring digital camera, the performance 
of the instrument is unstable and the orientation element is 
uncertain, so bring out the optical distortion error of aerial 
image. The camera focal length used in this system is fixed, 
so the distortion difference is the systematic error, which 
produces the same image for all the collected images. 
The inspection of camera can adopt the methods of 
optical laboratory inspection, laboratory inspection and 
on-duty inspection. At present, the main test method is 
laboratory test. The experimental field is composed of some 
mark points of known space coordinates. In the process of 
check and correct, the experimental field is photographed by 
the camera under check, and the internal azimuth elements 
and other elements affecting the shape of the beam are 
solved according to the method of single space resection or 
multi-space resection [2]. In 2D experimental campus, the 
system uses UAV digital camera Easy Calibrate to check and 
correct the digital camera SonyRx100), Table 1 shows the 
detection results and contents, with the origin of coordinates 
at the lower left corner of the image.  
TABLE I. THE CALIBRATION RESULTS OF SONY RX100 CAMERA 
Content of Check Calibration Value Remark 
x0 -0.008214mm 
the element of camera internal azimuth y0 -0.003216mm 
f(focal distance) 10.41234 
k1 (Radial distortion factor) 2.12E-10 
the coefficient of radial distortion 
k2(the factor of radial distortion) -8.14E-18 
p1 (the factor of eccentricity distortion) 3.14E-7 
Tangential distortion coefficient  
p2 (the factor of eccentricity distortion) -1.42E-7 
 
Image distortion correction -- indirect method. This 
method USES the coordinates on the corrected image to 
calculate the image coordinates of the corresponding points 
on the original image, and combines the gray interpolation 
method to realize image correction [3], As shown in figure 1. 
 
Figure 1. Image distortion correction schematic diagram 
International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
3 
After years of research, the calculate the deformation 
correction models getting thrown the error correction model), 
The model of error correction is:  


：  Image point coordinates x_0 and y_0 with 
image center as the origin: main point coordinates picture.   

：coefficient of radial distortion  
：coefficient of tangential distortion  
：Non-square scaling factor of pixels 
：The no orthogonal error coefficient of CCD array 
arrangement.  
The space resection method is used to calculate the 
coordinates of the camera in photography, which improves 
the precision of high external square elements and the 
precision of geometric calibration [4].  
The precise control of UAV attitude is mainly timely 
adjusted through the acquired signals of the attitude sensor, 
which generally includes two types: the angle sensor and the 
angular velocity sensor，The dip sensor is implemented 
indirectly by an acceleration sensor from three directions, 
The output signal values represent the current three axial 
acceleration values, If the UAV hovers in the air and stays 
still when the actual geological aerial measurement is made, 
then obtained acceleration value can be easily converted to 
obtain the real dip parameter [5]. It is impossible for a drone 
to remain stationary in the air for a long time in practical 
applications, When there is wind, the UAV may deviate 
from a certain direction when it is disturbed. Even if the 
UAV remains in a horizontal direction, the output value of 
the acceleration sensor will still deviate from the central 
value, resulting in the misjudgment value given by the 
control center. To avoid this kind of situation, often need to 
introduce three axis in the practical measuring angular 
velocity sensor and ultrasonic range finder, according to the 
three axis get up the acceleration and angular velocity and 
the Z axis direction real-time highly value the rate of change 
of the acceleration of X, Y axis direction, so as to draw close 
to actual information of the real Angle [5]. 
III. UAV IMAGE MATCHING 
In most cases, the UAV image mapping method adopts 
the image matching technology, which recognizes the 
eponymous point between two images or multiple images 
through corresponding matching algorithm. The common 
matching methods mainly include the following two 
categories: one is based on grayscale matching, the other is 
based on feature matching [4]. In the actual measurement in 
this paper, SIFT feature matching algorithm, which is most 
commonly used in feature matching mode, is adopted for 
high-precision matching of massive data. SIFT matching 
algorithm adopts the matching based on local feature values 
of the image [6]. This algorithm holds invariance for 
translation, rotation occlusion, etc. Therefore, it has strong 
stability in actual use. The feature matching process is shown 
in figure 2.  
 
Figure 2. Feature matching flow chart 
International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
4 
A. Pyramid image 
Pyramid image refers to the process of decomposing the 
original image to obtain a series of sub-images with different 
resolutions. These images are sorted from small to large 
according to the resolution, forming a group of overlapping 
pyramid-like images. The matching point is found in the top 
image, and the matching position is taken as the predicted 
position of the next layer. The matching result of this layer is 
taken as the initial matching position of the next layer, and 
then the matching is conducted successively. The matching 
result is used as control to match other feature points [7]. 
This process from top to bottom, from coarse to fine, ensures 
the reliability of image search process.  
In the pyramid image structure, the image is represented 
by hierarchical structure. At the top of the pyramid structure, 
data is stored at the lowest resolution possible, As the 
number of pyramid layers increases, The resolution of the 
data decreases successively. at the bottom of the pyramid, 
the data with the highest resolution can be stored to meet the 
needs of users [8]. In this way, different layers and different 
resolutions are adopted to store and display according to the 
needs of users, forming a pyramid structure with a higher 
resolution to a lower one and a smaller data volume to a 
larger one. This image pyramid structure is a typical 
hierarchical data structure used for image coding and 
progressive image transmission [9]. It is suitable for 
multi-resolution organization of raster data and image data 
and is also a lossy compression square of raster data or 
image data.  
B. Image feature point extraction 
Feature extraction refers to using a computer to present 
the image information of the same name in the image, which 
determines the common features in the image [7]. Image 
feature extraction generally depends on the distribution of 
gray in the image, and the position shape and size of features 
are determined through the information. 
SIFT feature matching algorithm consists of two parts 
[10]. The vector features are extracted from multiple 
images；SIFT is used to match feature vectors. Scale space 
representation is an expression based on region. Scale space 
is defined as the product of Gaussian convolution kernel and 
remote sensing image. Through the derivation of Koendetink 
and Babaud, it is proved that Gaussian kernel is the only 
linear kernel to realize scale transformation）[8]。 

In formula 4, L(x, y, σ) is the scale space,  is 
the Gaussian convolution kernel，  is the remote 
sensing image， x, y, σ are respectively represented by 
position parameters and scale parameters. The smaller the 
scale space factor is, the smaller the scale is.  

After defining the scale space, the scale space function 
can be used to build the Gaussian pyramid model, The scale 
proportion between two adjacent layers and the same rank 
pyramid affect the scale space function defined between two 
adjacent layers [11]. The scale between adjacent layers is 
defined as k, Define the scale factor as σ),  is the 
differential Gaussian pyramid function. At last, each sample 
point is compared with the adjacent points in the 
corresponding positions of the upper and lower scale space 
around the same scale space. If the detection point is local 
maximum or minimum, it is a candidate point of the image 
in this scale [12]. 
IV. THE EXPERIMENTAL TEST 
This experiment adopts the UAV model for the average 
Inspire V2.0 aerial vehicle four axis, the main parameters 
International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
5 
including the maximum altitude of 4500 meters, maximum 
rising speed of 5 m/s, the maximum level flight speed of 22 
m/s, maximum pitching Angle 35 ° 10 m/s wind power, 
aircraft cabin image sensor using Sony EXMOR 1/2.3)。The 
image device used in this experiment is Cannon5D Mark II, 
The image size is 36*24mm. Install Visual Studio 2017 on 
your laptop and configure Open CV for experimental testing, 
Using C++ to achieve SIFT image feature point extraction, 
The matching method is two dimensional feature point Brute 
Force Matcher, Set a certain threshold to filter the matching 
results, Using Find Homography function to set RANSAC 
method to eliminate error matching, SIFT was tested 
according to the above steps to understand its performance, 
and the performance was obtained as shown in table 2.  
TABLE II. FEATURES MATCHING PERFORMANCE TABLE 
 SIFT+BFMatch 
Image extraction point 33092/30012 
Time to generate the descriptor 178340ms 
Match the time 278731ms 
Threshold extraction points 16782 
Filter after mismatched points 9783 
 
The experimental results show that the method has a 
reasonable matching time. After filtering the threshold value 
and the basic matrix, the points basically cover the key area 
of the image. The pixel distribution is uniform with the low 
error. The system can meet the matching requirements, and 
the matching test image is shown in figure 3. The accuracy 
rate of the experimental matching is 91.63%. If the image is 
improved with light effect, the accuracy can reach 93.25%. 
Therefore, this research method has good anti-interference 
and high stability, and can be widely used in low-altitude 
image matching with interference factors.  
 
 
Figure 3. Feature Matching Experiments 
V. CONCLUSION 
Based on the preprocessing theory of aerial survey 
images, this paper studies the extraction method of image 
features and USES Visual C++ to realize SIFT extraction of 
image feature points. Brute Force Matcher, a 
two-dimensional feature point matching method, was 
selected for image region matching, Use the Find 
Homography function to set the RANSAC method to 
eliminate false matches, In this method, the matching 
accuracy was 91.63%~93.25% (according to the scene 
illumination) through the experiment of the geological and 
geomorphological images of a certain scene, and relatively 
satisfactory matching effect was obtained. However, due to 
the deficiencies of UAV itself, it is difficult to compare with 
professional image processing system. With the further 
development of communication technology and control 
technology, UAV will have a breakthrough application and 
development prospect in low-altitude measurement field in 
the future. 
 
REFERENCE 
[1] SONG Wenping. Research on Integration of Low Altitude Remote 
Sensing System of UAV and Related Questions of Image Date 
Processing[D].Xi’an: Chang’an University, 2016.06:38. 
[2] LI Yifei. Research on PID control in quad rotorcraft[J].Technology 
and Market, 2016.07:90. 
International Journal of Advanced Network, Monitoring and Controls        Volume 03, No.04, 2018 
6 
[3] WANG Jianxiong. Study of the key technology of low altitude 
phootogrammetry of unmanned aerial vehicle and the practice of 
large scale topography formation[D].Xi’an: Chang’an University, 
2011.12:54. 
[4] YU Huai, YANG Wen. A Fast Feature Extraction and Matching 
Algorithm for UVA[J].Journal of Electronic &Information 
Technology, 2016.03:510. 
[5] D. G. Lowe. Distinctive Image Features from Scale-Invariant Key 
points[J]. International Journal of Computer Vision, 2014, 60(2):263. 
[6] LI Xiang, WANG Yongjun, LI Zhi. Inter- TriadMisalignment of 
Vector Field Sensors in Attitude and Heading Reference Systems and 
Its Calibration[J]. Chinese Journal of Sensors and 
Actuators,2017.02:267-270 
[7] TAN Xiong, YU Xuchu, LIUJingzheng, HUANG Weijie. Object Fast 
Tracking Based on Unmanned Aerial Vehicle Video[C]// Proceedings 
of PACIIA, IEEE Presss, 2010:236. 
[8] WANG Donghua, YUE Dawei. Design and Implementation of large 
remote sensing image correction test system [J]. Computer 
Programming Skills & Maintenance,2015.12:118. 
[9] Liu Yawei. The review of UAV target detection and tracking 
method[J]. Aerodynamic Missile Journal, 2016,(9):54-56. 
[10] D. G. Lowe. Distinctive Image Features from Scale-Invariant Key 
points[J]. International Journal of Computer Vision.2014, 60(2). 
[11] CUI Zhe. Image’s characteristic point extraction and matching based 
on SIFT algorithm[D].Chengdu: University of Electronic Science and 
Technology of China. 2016.05:21-28. 
[12] LU Xiaopan. Empirical Study on the Mapping Precision Based on 
UAV Low-altitude Photogrammetry[D].Beijing: China University of 
Mining and Technology. 2014.06:53-56. 
 
 


English

Research and Realization on the key technology of Motion Image Matching Based on UAV

Shuping, Xu

Yiwei, Chen

Zhiyong, Zhang

Hao, Li

Implementation of Pan-Tilt System for Locating Based on ARM

Hejing, Wu

International Journal of Advanced Network, Monitoring and Controls          Volume 03, No.04, 2018 
92 
Application of AGC Technology in Software Radio 
Wu Hejing 
East University of Heilongjiang 
150086 
e-mail: wuhejing15@163.com 
 
 
Abstract—The characteristics of software radio are flexibility, 
openness, scalability. The hardware platform of software radio 
should be a general platform. This paper discusses automatic 
gain control(AGC) technology in software radio receiver and 
introduces an AGC algorithm applicable for DSP implement. 
This algorithm is tested in matlab and simulation results are 
provided. 
Keywords-Software Radio; Characteristics; AGC; Matlab 
I. INTRODUCTION 
At present, software radio technology is widely used in 
wireless communication, its basic idea is to use hardware as 
the basic platform of wireless communication. The A/D 
sampling data of signals are processed by various algorithms, 
and various communication functions are realized by means 
of software. This paper discusses the automatic gain control 
(AGC) algorithm in software radio. The function of AGC is 
to automatically adjust the gain of the amplifier according to 
the strength of the input signal received. Keep the output 
signal basically unchanged when the input signal changes in 
strength. An efficient AGC scheme suitable for DSP 
operation is proposed, which has fast convergence and 
accurate steady-state response characteristics. Moreover, it is 
simple to implement and can satisfy the needs of software 
radio system very well. 
II. CHARACTERISTICS OF SOFTWARE RADIO 
(1) flexibility. Software radio can be achieved by adding 
software modules. It's easy to add new features. It can 
communicate with any other radio station and act as radio 
frequency relay of other radio stations. 
(2) openness. Software radio adopts a standardized and 
modular structure. Its hardware can be updated or expanded 
with the development of devices and technologies. 
(3) scalability. Software radio can be upgraded by 
loading new software. 
III. ARCHITECTURE OF SOFTWARE RADIO 
Software radio architecture is the concrete design 
structure torealize the concept of software radio. It includes 
hardware, software and interface protocol. The design 
content must take into account the current situation and 
long-term development of wireless communication 
technology. Really unifies each standard. The structure of 
ideal software radio is mainly composed of antenna, RF front 
end, broadband A/D-D/A soft converter, General and special 
digital signal processors and various software components. 
The antenna of software radio generally covers a wide 
frequency band. For example, 1MHZ-2GHZ requires that the 
characteristics of each frequency band be uniform. To meet 
the needs of various businesses. The RF front-end mainly 
completes the tasks of up-conversion, filtering, power 
amplification and so on. Receiving realizes filtering, 
amplification, down conversion, and other functions. After 
digitizing the analog signal, the processing task is completely 
completed by the DSP software. In order to reduce the 
processing pressure of general AGC, A/D converter is 
usually used to transmit digital signals. After special digital 
signal processing devices, Reducing data flow rate, After the 
signal is changed to baseband, the data is controlled. 
DOI: 10.21307/ijanmc-2019-027
International Journal of Advanced Network, Monitoring and Controls          Volume 03, No.04, 2018 
93 
IV. SOFTWARE AND HARDWARE PLATFORM OF 
SOFTWARE RADIO 
The hardware platform of software radio should be a 
general platform. It may not have the highest efficiency for a 
particular communication system. But its openness and 
scalability enable it to adapt to a variety of wireless 
communication systems. And it can be flexible to add, 
subtract and modify. First, The hardware structure of 
software radio designed by people is basically a streaming 
structure. This structure is similar to the logical structure of 
wireless communication system. So its efficiency is higher. 
However, Because the direct coupling of each module in this 
structure is too close, There is a problem of pulling the 
trigger and moving the whole situation. Based on this 
situation, Bus architecture has been proposed. The bus 
structure has good openness. It is an ideal choice to realize 
software radio. There are many bus standards for industrial 
control bus. For example, ISA, PCI, EISA, VEM and so on. 
ISA bus and VME bus are widely used in current digital 
signal processing and industrial control. VEM bus has more 
advantages than ISA bus. The data of ISA bus is 16 bits. The 
address bus is 24 bits. Address space 16MB. Its bus 
bandwidth is only a few MHz. The data width of VME bus is 
32 bits. There are 32 address lines. The address space is 4GB. 
The bus bandwidth is tens of MHz. VEM bus is designed for 
multi-processor system. The ISA bus is a single processor 
bus. So VME bus is often used in software radio. 
Software radio operates in the so-called "software bus" 
mode. Its software structure consists of software modules 
with standard interfaces. Each software module defines 
different application functions. The integrated operation can 
be achieved by inserting the bus. Realize the complete 
communication function. The integrated operation mode of 
"Software Bus" requires the unification of software interface 
standards. Improve software openness and reusability. 
Realize online software sharing in multiple environments. 
Applications based on JAVA technology can work on 
various platform systems. JAVA only needs to be written 
once. It can run anywhere. It is easy to implement function 
expansion and module embedding. It can realize the rapid 
opening of new business. JAVA plays an important role in 
realizing software reusability. 
V. PRINCIPLE OF AUTOMATIC GAIN CONTROL BY 
SOFTWARE 
The block diagram of the principle of automatic gain 
control is shown in Figure 1.The IF sampling signal x(n) is 
amplified by a controllable gain amplifier. The gain is A 
(n).The output signal y(n)= A(n)x(n),Calculate the 
logarithmic value of the level of signal y(n).Compared with 
reference level log(R),An error level e(n) is generated. The 
gain of the amplifier is continuously adjusted by negative 
feedback with e(n).The output log (y (n) is gradually 
approached to the reference level log (R) until the circuit 
reaches equilibrium. The parameter alpha controls the 
adjustment time of AGC circuit, which is a constant related 
to time. The logarithmic and exponential operation of signal 
level in circuit is to make the constant of adjusting time of 
control circuit independent of input level. The mathematical 
analysis of the block diagram is as follows: 
 
Figure 1. AGC algorithm implementation principle1 
International Journal of Advanced Network, Monitoring and Controls          Volume 03, No.04, 2018 
94 
The recurrence formula of magnification factor is as 
follows: 
aRnxanAnxnARanAnA  ])(1)[(])()([)()1(
When the amplitude after passing through the IF amplifier is 
less than R,
)()( nxnAR 
 as positive, The amplification 
factor increases.
)(ny
 increases, Similarly, when the 
magnitude is greater than R,
)()( nxnAR 
 as negative. 
The magnification decreases. 
)(ny
reduce, thus small 
signal amplification can be realized. Large signal attenuation, 
the signal is controlled within a certain range. Suppose the 
input signal is 
)()( ncunx 
 
aRacnAnA  ]1)[()1(
 
)(])1(1[)( nuac
c
R
nA n
 
The gain after reaching a fixed state is c
R
，Time 
constant is ac
1
 
The schematic diagram of AGC's second scheme is as 
follows: 
 
 
Figure 2. AGC algorithm implementation principle2 
 
The recursive formula is: 
}])()(log{}[log{)}(log{)}1(log{ nxnARanAnA 
In order to make the response time faster, Before comparing 
with the comparative value R, The output value of the 
amplifier 
)()()( nxnAny 
 is logarithmic before 
comparison. When the magnitude is less than R, 
})()(log{ nxnA
 is less than
}log{R
.Therefore, 
})()(log{}log{ nxnAR 
 is positive. The 
magnification becomes larger, Thus, the small signal 
is enlarged. When the magnitude is greater than R, 
})()(log{ nxnA
 is larger than 
}log{R
, therefore 
})()(log{}log{ nxnAR 
 is negative, The 
amplification factor 
)(nA
 decreases, thus the attenuation 
of large signal is realized. Through the above algorithm, the 
small signal is enlarged, Large signal attenuation. Suppose 
the input signal is 
)()( ncunx  ，have  
}/log{]1)}[(log{)}1(log{ RcaanAnA 
 
International Journal of Advanced Network, Monitoring and Controls          Volume 03, No.04, 2018 
95 
)(])1(1}[log{)}(log{ nua
R
c
nA n
 
As can be seen from the above formula, The gain of the 
stabilized signal is c
R
，The time constant is approximately 
equal to a
1
 
VI. SIMULATION RESULTS 
The AGC algorithm is implemented by matlab. Two 
AGC algorithms are added to the demodulation of a 
narrowband QPSK signal. By observing and comparing the 
AGC output signals, The two schemes are proved to be 
completely feasible. The magnification becomes larger, Thus, 
the small signal is enlarged. When the magnitude is greater 
than R,
})()(log{ nxnA
is larger than 
}log{R ，
therefore 
})()(log{}log{ nxnAR 
is negative, The 
amplification factor
)(nA
decreases, thus the attenuation of 
large signal is realized.Through the above algorithm,The 
simulation results of two AGC schemes are shown in Fig. 3 
and Fig. 4 respectively. 
 
 
Figure 3. The simulation results of AGC algorithm when the input signal of the first AGC scheme change 
0 1000 2000 3000 4000 5000 6000 7000 8000
-5
-4
-3
-2
-1
0
1
2
3
4
5
0 1000 2000 3000 4000 5000 6000 7000 8000
-30
-25
-20
-15
-10
-5
0
5
10
15
0 1000 2000 3000 4000 5000 6000 7000 8000
0
1
2
3
4
5
6
7
International Journal of Advanced Network, Monitoring and Controls          Volume 03, No.04, 2018 
96 
 
 
Figure 4. The simulation results of AGC algorithm when the input signal of the second AGC scheme changes 
VII. CONCLUDING REMARKS 
Since the introduction of software radio,its brand-new 
software value concept solves the problem of wireless 
standard interoperability.，Then it triggered a worldwide 
research climax. Software radio receivers, Whether the 
signal amplitude is stable or not will affect the timing 
synchronization and phase synchronization recovery of the 
receiver. This will determine whether the signal can be 
received with high quality, so automatic gain control (AGC) 
is applied to the input signal. High quality signal reception 
can be achieved. These two AGC schemes have fast 
convergence and accurate steady-state response 
characteristics, and the algorithm is simple.It is convenient 
for software radio system to be realized by DSP. 
 
REFERENCE 
[1] Research on the Practical Course "Software Radio System Design 
and Verification" [J]. DuanRui, Chen Zhuming, Zou Lin. 
Experimental Science and Technology. 2016 (06)  
[2] Research on SOPC-based Software Radio System [J]. Yang Zhengyu, 
Li Bing. Microcomputer and Application. 2010 (01) 
[3] Application Research of Radar System Based on Software Radio 
Technology [J]. Liu Xiaoping. Information and Computer 
(Theoretical Edition). 2018 (21)  
[4] Application of Software Radio Technology in FM Transmitter Design 
[J]. Sandik.Digital Communication World. 2017 (02)  
[5] Application of Software Radio Technology in Ship Communication 
[J]. Wang Huan. Strait Science and Technology and Industry. 2017 
(09) 
[6] Application and Development of Software Radio Technology [J]. 
Wang Hongbo. Computer Programming Skills and Maintenance. 
2016 (19) 
0 1000 2000 3000 4000 5000 6000 7000 8000
-5
-4
-3
-2
-1
0
1
2
3
4
5
0 1000 2000 3000 4000 5000 6000 7000 8000
-40
-30
-20
-10
0
10
20
30
0 1000 2000 3000 4000 5000 6000 7000 8000
0
1
2
3
4
5
6
7
8
9
Fig. a Fig. b 


Application of AGC Technology in Software Radio

Basu, D.

Basu, S.

Reddy, M.

Gupta, K.

Chandy, M.

Clinical and laboratory profile of anti-M

MILOŚEVIĆ, Predrag

MILOŚEVIĆ, Vladimir

MILOŚEVIĆ, Grigor

TRADITION AND INNOVATION IN SERBIAN 21ST CENTURY ARCHITECTURE – TRADITIONAL MATERIALS, MODERN FORMS AND SUGGESTED DESIGNS

Williams, Rachel L.

Art and contemporary Irish emigrants: Aesthetics and networks of affect in David Monahan’s ‘Leaving Dublin’

Siergiej, Małgorzata

Sudoł-Szopińska, Iwona

Zwoliński, Jerzy

Śladowska-Zwolińska, Anna Maria

Role of intrapartum ultrasound in modern obstetrics – current perspectives and literature review

Jiang, Kun

Wang, Jianguo

Research on Improved Adaptive ViBe Algorithm For Vehicle Detection

Clem, Peter

 
THE PLACE OF PSYCHOMETRIC TESTING IN PILOT 
SELECTION 
Peter Clem 
BSc, DipMilAvn, PostGradDipSci, MBBS, 
PostGradDipAvMed, RACGP 
 
Corresponding author:  
Peter Clem 
peterclem.feedback@gmail.com 
 
ABSTRACT 
Aim 
Pilot training has always been a relatively 
expensive undertaking. So attempts to control 
these costs by predicting the likelihood of 
success or failure is a constant that is almost 
as old as aviation itself. Incorporation of 
Psychometric testing was made to pilot 
selection in the years between the first and 
second world wars. Despite the many changes 
that have occurred in this area, psychometric 
testing continues to feature in modern systems 
of pilot aptitude testing. This paper reviews 
some of the history of psychometric testing in 
pilot selection. 
 
BACKGROUND 
After World War One efforts to predict success in 
pilot training were occurring both in Europe and 
the United States. In Europe, the French and 
British approaches tended to focus on the 
Physiological challenges of aviation. In the United 
States, the approach tended to focus more on the 
psychological difficulties. This dichotomous 
approach continued well into the Second World 
War. It was not until the landmark Pensacola 1000 
study in the United States in 1945 that 
demonstrated the superiority of the Psychological 
approach. (1) 
900 US Navy flight training cadets were subjected 
to 60 different psychological, psychomotor and 
physiological tests. The Pensacola 1000 study 
determined that the physiological tests were not 
predictive of success more than chance. The 
study concluded that psychometric and 
psychomotor tests were predictive of success on 
flight training. 
The Pensacola 1000 study became the model for 
pilot psychometric testing from 1945 until the 
present day. The research led to the creation of 
the Naval Aviator Test Battery. The Naval Aviator 
Test Battery included the Wonderlic Personnel 
Test (a test of general ability or intelligence), the 
Bennett Mechanical Comprehension Test (a test 
of mechanical interest and skills), and the Purdue 
Biographical Inventory (a measure of morale, 
interest, and attitudes). 
 
PSYCHOMOTOR TESTING 
The Pensacola 1000 study demonstrated that 
psychomotor tests were predictive of success in 
pilot training. Despite this result, Psychomotor 
tests had problems with ease of use, reliability, 
and standardisation. As a result, psychomotor 
tests were omitted from the Naval Aviator Test 
Battery. Psychomotor tests fell out of use in the 
United States in the decade after the Second 
World War. (1) 
Outside of the United States, the experience was 
different. The Royal Air Force (RAF) and the Royal 
Australian Air Force (RAAF) persisted with 
electromechanical psychomotor tests from the 
1940s until well into the 1990s.  
Pilot aptitude testing has also included various 
combinations of other measures. These have 
included previous flight experience, previous 
service experience, interview results, performance 
on work sampling, and results at flight screening. 
This has been an attempt to boost the relatively 
modest predictive power of psychometric and 
psychomotor testing. (2, 3) 
 
DEFENCE MECHANISM TEST 
In the 1970s and 1980s, Scandinavian countries 
developed the Defence Mechanism Test (DMT), 
which was a significant departure from the 
approach that had developed following the 
Pensacola 1000 study. The DMT is a projective 
test based on assumedly anxiety-provoking 
images. The images are exposed through a 
tachistoscope. The tachistoscope gradually 
increases the image exposure from 5 to 2000 
milliseconds. The rationale for the test as a 
 
selection instrument for stressful occupations is 
that psychological defences bind psychic energy 
necessary for coping with stressful situations. 
Furthermore, those subjects with maladaptive 
strategies for dealing with stress will perform 
worse on the test. The Scandinavians reported 
significant predictive ability for the DMT. This 
approach was trialled by Air Forces outside of 
Scandinavia, including the RAAF. The DMT failed 
to demonstrate the same results. A study by 
Ekehammar et al. in 2005 aimed to understand 
why this was so. They concluded that the DMT 
does not measure what it purports to measure. 
They found that a more plausible explanation was 
that DMT performance reflects information 
processing difficulty due to anticipatory or test 
anxiety. (4) 
 
COMPUTER TECHNOLOGY 
Since the 1980s computer technology has been 
introduced into pilot aptitude testing. Various 
systems have been deployed which combine 
psychometric and psychomotor tests into the one 
device. (1) It is important to realise that these 
machines are based on electronic versions of pre-
existing psychometric and psychomotor tests. 
Because of this, Bartram et al. 1995 concluded 
that this technology is not expected to significantly 
enhance the prediction of success or failure on 
pilot training. (5) 
A strength of computer-based devices is that the 
Psychomotor element doesn't suffer from the 
issues that plagued electromechanical devices of 
the past. This led to the United States military 
reintroducing psychomotor testing into their test 
batteries in the 1980s. Computer-based systems 
have thus enabled reintroduction and combination 
of psychometric and psychomotor testing into a 
single device. The widespread uptake of these 
devices by military and non-military users around 
the globe has helped to standardise the 
assessment process. 
 
STATISTICAL ISSUES 
Although there has been widespread and 
continuous use of psychometric testing over a very 
long period, the predictive abilities of these tests 
have always been modest.  
A landmark meta-analysis study by Hunter and 
Burke in 1992 reported validity coefficients as a 
function of predictor type (table 1). (6) 
 
Predictor Measure Mean r 
General Cognitive 0.1924 
Personality 0.1168 
Information Processing 0.2256 
Job Sample 0.3272 
Biographical Inventory 0.2646 
Psychomotor Co-ordination 0.3035 
Composite Battery 0.1934 
Other 0.0889 
Total 0.1973 
 
Table 1. Validity Coefficients as a Function of 
Predictor Type 
 
The researchers reported that in general, job 
sample measures were the best predictors of 
performance, followed by psychomotor co-
ordination and biographical inventories. 
Somewhat depressingly, Hunter and Burke 
reported that the analysis showed a decline in the 
mean validity correlations over the previous 50 
years. 
Another disappointing finding was that for the 
personality measures (mean correlation of 0.1168) 
the 95% confidence interval was +/- 0.2644. An 
interval which includes zero meaning that this 
measure is no more predictive than a coin toss. (6) 
 
SO WHY ARE THESE TESTS NOT ABLE TO BE MORE 
PREDICTIVE? 
In a 1996 paper by Damos et al, Pilot Selection 
Batteries: Shortcomings and Perspectives, the 
authors noted the low correlation between 
predictors and outcome criteria as described by 
Hunter and Burke. They noted that predictive 
validities based on intelligence tests and 
personality tests were in the range of 0.15 to 0.20. 
Damos noted that selection batteries that 
combined intelligence, psychomotor, personality 
and information processing tests could achieve 
predictive validities in the range of 0.20 to 0.40. (7) 
Damos offered the following list of potential 
explanations for why these tests aren’t more 
predictive: (7) 
1. Sudden workforce changes. Leading to 
military altering criteria for pass-fail and thereby 
adversely affecting the correlations. 
 
2. The use of pass-fail criteria. Dichotomising 
the criterion variable at the mean results in a 38% 
reduction of effective sample size when the 
correlation is between 0.20 and 0.50. A high 
success rate in pilot training effectively limits the 
biserial point correlation between the predictor and 
the criterion variable. 
3. Test development. Historically the tests were 
not based on task analysis but were assumed a 
priori to have some validity for predicting success 
in pilot training. Many of the tests were based on 
psychological theories of human cognition and 
personality which may or may not play a significant 
role in performing such a complex task as flying an 
aircraft. 
 
SO WHY DO MILITARIES AND AIRLINES CONTINUE TO 
UNDERTAKE THESE TYPES OF TESTS? 
Although psychometric tests are unable to provide 
significant prediction in isolation, when they are 
combined into selection batteries, they provide 
increments in prediction that continue to be 
attractive to organisations that are responsible for 
candidate selection. (2, 3) A common and 
interesting observation is that correlations with 
success on pilot training are not reflected in 
success on operational training. (7) 
As long as pilot training continues to be expensive 
and while there is a large number of applicants for 
a small number of training places it is likely that 
this approach will continue, despite its limitations. 
On the other hand, if there becomes a severe 
shortage of pilots (as predicted by ICAO), the 
limitations of this approach may become more 
apparent. 
 
AIRLINE INVOLVEMENT 
Many airlines around the world have incorporated 
psychometric testing into their selection 
processes. Somewhat paradoxically, United 
States airlines have not employed these tests as 
much as many overseas Airlines, due in part to the 
particular regulatory framework in which they 
operate. (7) 
The International Aviation Transport Association 
(IATA) have published Guidance Material and 
Best Practices for Pilot Aptitude Testing. (8)  
These guidelines make the following claims for 
pilot aptitude testing;  
• “If correctly implemented, a pilot aptitude 
testing system can contribute to considerable 
cost savings for the airline as well as:  
• Enhanced safety  
• Decreased training costs 
• Increased training and operational 
performance success rates 
• More positive working environment 
• Reductions in labor turnover 
• Enhancement of the flight operations 
department and airline’s brand” 
The guidelines are based on a large survey of the 
practices of member airlines. Although the 
guidelines make significant claims for the benefit 
of pilot aptitude testing, they do not address some 
of the limitations identified in research journals 
previously identified in this paper. The guidelines 
do note the following; 
“Aptitude testing systems are not "perfect" in 
predicting the future performance of pilots. 
However, if they are developed and designed 
responsibly, they can offer valuable guidance to 
the operator. There is consensus amongst experts 
in the field of aptitude testing that performance of 
pilots can be reasonable well predicted employing 
testing. Opinions differ on a) how long the 
predictions are valid, b) which category of 
performance can be predicted best and c) how 
detailed the prediction can be”. (8) 
As previously noted, research has indicated that 
correlations with success on pilot training are not 
reflected in success on operational training. So it’s 
worthwhile to consider to whom the airlines are 
applying these IATA guidelines. If they are applied 
to airline cadets then based on the military 
experience there they might predict success with 
training.  
If the test batteries are being applied to pilots being 
recruited from the military, or other airlines, then 
this expectation is probably not realistic. 
Alternatively, airlines may expect that (for qualified 
pilots) the batteries may select candidates who are 
‘safer’ or more compatible with the ‘culture’ of the 
company. The evidence base for this expectation 
is not currently well described in the scientific 
literature. 
 
THE FUTURE 
Since the 1980s there has been a convergence of 
psychometric and psychomotor testing in terms of 
their incorporation into computer-based devices. 
 
At the same time computer technology has been 
incorporated into the cockpit with resulting 
automation of critical roles. As these processes 
continues the screening device, the simulator and 
the aircraft may converge to the point that they are 
largely identical from the view point of the “pilot”. 
At this stage very little if any selection or training 
will take place in real aircraft. 
A survey by financial services firm UBS estimated 
that moving from 2 pilots to 1 pilot for airline 
operations would yield a potential profit of $15 
billion. The study also noted that 70-80% of 
accidents are the result of human error and that 
15-20% of those are due to crew fatigue. (10) It is 
likely that these drivers will result in greater 
automation of the cockpit to the point where a 
pilots role may be unrecognisable from what it is 
today. The level of automation may result in pilot’s 
tasks becoming more similar to that of a Main 
Control Room (MCR) operator in a nuclear power 
plant. 
A study by Zhang et al. looked at any correlation 
between a Psychometric measure known as 
general mental ability (GMA) and the performance 
and safety compliance of main control room 
(MCR) operators in nuclear power plants. The 
study noted that GMA is the best single predictor 
of work performance with the criterion related 
validity as high as .51. (9) 
In this context, it’s interesting to consider that a 
change in the “task” might be the missing 
ingredient which finally delivers on the promise of 
psychometric testing. 
BIBLIOGRAPHY 
1. Griffin GR, Koonce JM. Review of 
psychomotor skills in pilot selection research 
of the U.S. military services. Int J Aviat 
Psychol. 1996;6(2):125-47. 
2. Carretta T. Pilot Candidate Selection Method: 
Still an Effective Predictor of US Air Force 
Pilot Training Performance. Aviation 
Psychology and Applied Human Factors. 
2017. 
3. Weissmuller JJ, Damos DL. Improving the 
Pilot Selection System: Statistical Approaches 
and Selection Processes. The International 
Journal of Aviation Psychology. 2014. 
4. Ekehammar B, Zuber I, Konstenius ML. An 
empirical look at the Defense Mechanism Test 
(DMT): reliability and construct validity. 
Scandinavian journal of psychology. 
2005;46(3):285-96. 
5. Bartram D. Validation of the Micropat Battery. 
International Journal of Selection and 
Assessment. 1995;3(2):84-95. 
6. Hunter D. Predicting Aircraft Pilot- Training 
Success: A Meta-Analysis of Published 
Research. - 
ijap_meta_analysis_of_pilot_selection.pdf. 
1992. 
7. Damos DL. Pilot selection batteries: 
shortcomings and perspectives. Int J Aviat 
Psychol. 1996;6(2):199-209. 
8. Association IAT. Guidance Material and Best 
Practice for Pilot Aptitude Testing. Montreal 
Geneva: International Air Transport 
Association; 2012. 
9. Zhang J, Yongjuan L, Changxu W. The 
Influence of Individual and Team Cognitive 
Ability on Operator’s Task and Safety 
Performance: A Multilevel Field Study in 
Nuclear Power Plants. 
https://journals.plos.org/plosone/article?id=10.
1371/journal.pone.0084528 
10. Would you get on a pilotless plane? BBC 8 
August 2017. 
https://www.bbc.com/news/business-
40860911 


THE PLACE OF PSYCHOMETRIC TESTING IN PILOT SELECTION

Nichols, Linda

The Circle of Willis, Aneurysms and Subarachnoid Haemorrhages: a Historical Narrative of Parallels from Observation to Intervention

Not a member yet