We are IntechOpen,

the world’s leading publisher of

Open Access books
Built by scientists, for scientists

4,700
Open access books available
121,000
International authors and editors
135M
Downloads

Our authors are among the

154
Countries delivered to
TOP 1%
most cited scientists
12.2%
Contributors from top 500 universities

Selection of our books indexed in the Book Citation Index

in Web of Science™ Core Collection (BKCI)

Interested in publishing with us?

Contact [email protected]
Numbers displayed above are based on latest data collected.
For more information visit www.intechopen.com
 Chapter 12

Kalman Filter for Moving Object Tracking: Performance

Analysis and Filter Design

Kenshi Saho

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/intechopen.71731

Abstract
This chapter presents Kalman filters for tracking moving objects and their efficient
design strategy based on steady-state performance analysis. First, a dynamic/measure-
ment model is defined for the tracking systems, assuming both position-only and
position-velocity measurements. Then, problems with the Kalman filter design in track-
ing systems are summarized, and an efficient steady-state performance index proposed
by the author [termed the root-mean-squared error index (the RMS index)] is introduced
to resolve these concerns. The analytical relationship between the proposed RMS index
and the covariance matrix of the process noise is shown, leading to a proposed design
strategy that is based on this relationship. Theoretical performance analysis is conducted
using the performance indices to show the optimality of the design strategy. Numerical
simulations show the validity of the theoretical analyses and effectiveness of the pro-
posed strategy in realistic situations. In addition, the optimal performance of the
position-only-measured and position-velocity-measured systems is analyzed and com-
pared. This comparison shows that the position-velocity-measured Kalman filter track-
ing is accurate when compared with the position-only-measured filter.

Keywords: Kalman tracking filter, moving object tracking, steady-state analysis,

performance index, filter design, process noise

1. Introduction

Remote monitoring systems for cars and robots require accurate tracking of moving objects.
Representative tracking algorithms include the Kalman filter [1–5] and its variants, such as the
extended/unscented Kalman [6–9] and particle filters [10–12]. These can accurately track
movement based on adaptive filtering by using a state-space model.
To use the Kalman filter for the tracking of moving objects, it is necessary to design a dynamic
model of target motion. The most common dynamic model is a constant velocity (CV) model

© 2018 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
234 Kalman Filters - Theory for Advanced Applications

[1, 10], which assumes that the velocity is constant during a sampling interval. This model has
been used in many applications because of its versatility, effectiveness, and simplicity. How-
ever, in almost conventional tracking systems, the selection of process noise (zero-mean white
noise in the dynamic model) is conducted empirically [4, 6, 8]. This is because conventional
studies tend to assume that process noise takes one of a limited number of forms, which is
known as appropriate selections. Thus, despite the large number of investigations into Kalman
filter trackers, the optimal selection of a process noise model has not been discussed. The
general problems of model selection for Kalman filter trackers were discussed by Ekstrand in
2012 [1]. In the years since, further research on these issues has been conducted, but no
satisfactory solutions to the abovementioned problems have been presented. Crouse [13]
described a general solution for optimal trackers in a steady state. However, this method also
requires an empirical selection of the dynamic models. A detailed analysis of the Kalman filter
has been provided for various applications, including global navigation satellite systems [14]
and video trackers [15]. However, only limited systems have yet been considered, and no
definitive parameter-setting procedure for the Kalman tracking filter has been provided.
Although various criteria have been proposed and investigated for the design of Kalman filters
and its variants to achieve better tracking accuracy, robustness, and real-time capability, rela-
tionship between these performance indices and the model parameters such as the process
noise variance is not discussed even in recent studies [16].
Another significant problem in a measurement model of the conventional Kalman tracking
filter is that most studies consider only position measurements and therefore cannot make full
use of modern sensors that are able to measure velocity, such as ultrawideband Doppler radar
[17, 18]. Moreover, sensor fusion based on Internet of Things technology also enables the
simultaneous measurement of position and velocity (e.g., sensor data fusion based on the
communication between radars/lasers/sonars and speedometers embedded in targets). Conse-
quently, Kalman filters for such systems have become an important area of research [19–24]. In
Ref. [24], the extended Kalman filter for radar measurements is modified for range (position)
and range-rate (velocity) measurements, and its effectiveness in realistic radar applications is
verified. However, concrete design criterion is not shown. The number of conventional studies
on position-velocity-measured (PVM) Kalman filters is smaller than those on the more com-
mon position-only-measured (POM) Kalman filters, and the performance and design of PVM
Kalman filters are not sufficiently considered.
To resolve the two problems described above concerning the process noise selection and PVM
systems, our previous work clarified the fundamental properties of PVM tracking filters [25,
26] and generated an efficient performance index to design an optimal process noise matrix [3,
5]. In the studies of PVM tracking filters [25, 26], fixed-gain PVM filter properties were
analytically clarified, but there was no optimization of the PVM Kalman filters. In our work
on the process noise matrix [3], an optimal POM Kalman filter, with respect to position
prediction, was presented. In this chapter, an appropriate process noise design strategy, based
on our proposed efficient steady-state performance index (introduced in Section 3), and its
applicability are verified. Our previous work highlighted the following issues, which we
address in this chapter:
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 235
http://dx.doi.org/10.5772/intechopen.71731

i. Analysis of the performance of a PVM Kalman filter with a CV model, based on the
proposed index.

ii. Application of the proposed process noise design strategy to a PVM Kalman filter.
iii. Comparison of the performance of optimal POM and PVM Kalman filters.
This chapter presents the theoretical analyses and simulations required to tackle these issues.
The remainder of this chapter is organized as follows: Section 2 defines the tracking filtering
problem dealt in this chapter and explains the existing concerns and models for POM and PVM
Kalman filter design. Section 3 introduces our proposed efficient performance indices with their
mathematical formulations. Section 4 presents the proposed process noise design strategy based
on the performance index. Section 5 shows the theoretical analysis of the optimal POM and
PVM Kalman filter performance in a steady state. The effectiveness of the PVM Kalman filter is
proven by the comparison with the POM filter. Section 6 shows realistic maneuvering-target-
tracking application examples. Section 7 concludes this chapter and proposes future tasks.

2. Problem statement

This section introduces the Kalman filter for moving object tracking and defines the model
assumed in this chapter.

2.1. Dynamic model

The Kalman filter for tracking moving objects estimates a state vector comprising the param-
eters of the target, such as position and velocity, based on a dynamic/measurement model. For
simplicity, this chapter deals with a typical second-order one-dimensional Kalman filter
tracker whose true state vector is defined as

xt ¼ ðxt vt ÞT , (1)

where xt and vt are the true position and velocity of the target moving object, respectively,
and T denotes the transpose. The assumed dynamic model is a CV model, which is a simple
and popular model for tracking moving objects. The CV model assumes that the velocity is
constant during the sampling interval, which is expressed as

xtk ¼ Φxtk 1 þ wk , (2)

where xtk denotes the true state at time kT, T is the sampling interval, wk is the process noise with
covariance matrix Q, and Φ is the transition matrix from kT to ðk þ 1ÞT, which is expressed as

1 T
 
Φ¼ , (3)
0 1

The Kalman filter predicts the target state based on this dynamic model.
236 Kalman Filters - Theory for Advanced Applications

2.2. Measurement model

The measurements are simply modeled as

zk ¼ Hxtk þ vk , (4)

where zk denotes the measurement vector, H denotes the measurement matrix, and vk is the
measurement noise with covariance matrix R. This chapter considers two types of measure-
ment systems, which are discussed as follows.

2.2.1. Position-only-measured system

The POM system assumes that the sensors (such as radar, laser, and sonar) can measure only
the position of the target. This is a general assumption in the moving object tracking. H and R
of this model are expressed as

H ¼ ð 1 0 Þ, (5)

R ¼ ðBx Þ, (6)

where Bx is the variance of the position measurement errors.

2.2.2. Position-velocity-measured system

The PVM system assumes that the sensor system can measure position and velocity simulta-
neously. One example of the PVM model system is a pulse Doppler radar. Sensor fusion
systems using communications of position/velocity sensors can also be expressed by the PVM
model. H of this model is expressed as
!
1 0
H¼ : (7)
0 1

We now assume that the noises of position and velocity measurements are uncorrelated, and R
of PVM systems under this assumption is defined as
!
Bx 0
R¼ : (8)
0 Bv

where Bv is the variance of the velocity measurement errors.

2.3. Kalman filter tracking

The Kalman filter tracker based on the abovementioned models sequentially estimates state
vectors via the Kalman filter equations. The prediction and estimation are calculated as
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 237
http://dx.doi.org/10.5772/intechopen.71731

x~ k ¼ Φb
xk 1, (9)

x k ¼ x~k þ Kk ðzk
b x k Þ,
H~ (10)

where predicts and estimates are denoted by ~ and ^, respectively, and Kk denotes the Kalman
gain that minimizes the errors in the estimated position and velocity. Kk is calculated as


~ k H T HP
Kk ¼ P ~ k HT þ R : (11)

where Pk is the covariance matrix of errors determined from

P b k 1 ΦT þ Q:
~ k ¼ ΦP (12)

bk ¼ P
P ~k ~k:
Kk HP (13)

2.4. Aspects of tracking filter design

Moving object tracking obtains accurate and sequential estimation of the target position and
velocity by using Eqs. (9)–(13). As indicated in Eqs. (1)–(13), the design parameters of the
Kalman filter tracker are elements of the covariance matrix of the process noise Q. We must
set Q to achieve tracking errors that are as small as possible. Thus, we must know how to
design an appropriate Q. Moreover, we must be able to define the evaluation index of the filter
performance. However, these issues have not been sufficiently deliberated because the selec-
tion of Q has not been sufficiently addressed in previous studies. Here, the design of Q is
empirically carried out.
In the conventional tracking systems, the most commonly used random acceleration (RA)
process noise is often selected because it has a better performance. Its Q is
!
T 4 =4 T 3 =2
Qra ¼ σ2q : (14)
T 3 =2 T 2

The appropriate selection of σq is important because σq (and sensor noise variance R) directly
determines the performance of the tracking filter with the CV model. However, in conven-
tional studies, process noises and their parameters are empirically selected, and the validity of
the selection is discussed only casually [1, 16]. Many conventional tracking systems select the
RA process noises (Qra ), with variance σq set based on the assumed target motion. However,
no definitive method of determining σq has been established. Although tracking index defined
by Kalata [27] is known as an effective design parameter, its empirical selection is still required.
Moreover, the validity in selecting the RA process noise is also questionable. Various other
forms of Q are known and have been used for different target motions [12]. For example,
random velocity model [2] and the diagonal Q, which do not include correlations in process
noise [7], are also frequently used. However, for the reasons discussed earlier, the differences
in performance between the various process noise models are not known.
238 Kalman Filters - Theory for Advanced Applications

3. The efficient steady-state performance index (RMS index)

The process noise selection problems discussed in Section 2.3 must be solved to effectively
design Kalman tracking filters. Thus, we must properly evaluate the performance of the filter.
The effective steady-state performance index was derived [3] and is termed root-mean-squared
error index (an RMS index). This section introduces the RMS index for POM and PVM systems
and shows the analytical relationships between the RMS index and Q.

3.1. Definition of RMS index

In tracking filtering, the following two functions are required:

• Function 1. Reduces random errors caused by measurement noises.
• Function 2. Tracks targets with complicated motions (e.g., accurate tracking of an acceler-
ating target is required for the CV model).

The RMS index is proposed for the comprehensive evaluation of the performance of these two
functions and is defined as
rffiffiffihffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiiffiffi
εp  E ðxtak x~ k Þ2 (15)

where x~ k is the predicted target position (second element of x~k ), E[] indicates the mean with
respect to k, and xtak is the true position of a constant acceleration target which is

xtak ¼ ac ðkT Þ2 =2 (16)

where ac is constant acceleration of the target. In the Kalman filter tracker using the CV model,
it is assumed that the target velocity is constant during the sampling interval. Thus, for the
constant acceleration target, a steady-state bias error occurs because of the difference between
the target motion and the assumed dynamic model. Moreover, x~ k includes random errors due
to measurement noise. Thus, the RMS index εp expresses both bias errors and random errors.
With the steady-state bias error due to the model/motion difference of eac and the steady-state
standard deviation of the random errors in x~k of σp , εp is expressed as
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
εp ¼ e2ac þ σ2p (17)

σp expresses the performance corresponding to Function 1 and eac expresses the performance
corresponding to Function 2. The smaller these errors are, the better is the tracking filter. Thus,
the minimum εp achieves the best tracking filter in a steady state.

3.2. RMS index of a POM system

One important advantage of the RMS index is that it can be expressed in closed form. The
closed form of εp for the POM system was derived in Ref. [3]. This subsection introduces the
RMS index and its relationship to the design parameter Q in the POM system.
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 239
http://dx.doi.org/10.5772/intechopen.71731

First, the arbitrary Q is defined as

!
a b
Qgen ¼ (18)
b c

where a > 0, b > 0, and c > 0, and the dimensions of a, b, and c are [m2], [m2/s], and [m2/s2],
 
respectively. For example, substituting ða; b; cÞ ¼ σ2q T 4 =4; σ2q T 3 =2; σ2q T 2 into Eq. (18) gives the
Qra of Eq. (14) and b = 0 leads to the diagonal Q. The analytical relationship between Qgen and
εp is expressed by the following closed form.

ac 2 T 4 2α2 þ 2β þ αβ
ε2p, pom ¼ þ   Bx (19)
β2 α 4 2α β

 T
where α and β are components of the steady-state Kalman gain K∞ ¼ α; β=T calculated
from (a, b, c) using the following equations:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
C2 ð16 þ 4A 4B þ CÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Cþ Cð16 þ 4A 4B þ CÞ Cð2A 2B þ C
β¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ (20)
4 8 Cð16 þ 4A 4B þ CÞ 8

A ¼ a=Bx , B ¼ bT=Bx , C ¼ cT 2 =Bx (21)

α¼1 β2 =C (22)

The derivation process of these equations is shown in Ref. [3]. As shown in Eqs. (19)–(22), the
optimal (a, b, c) is designed minimizing εp .

3.3. RMS index of a PVM system

In a similar manner to the treatment of the POM system, this subsection introduces the RMS
index of a PVM system and its relationship to Qgen . The RMS index of the PVM system is
!2
g2 α; β; η; θ Bx þ g3 α; β; η; θ T 2 Bv
   
2 2η θ
ε2p, pvm ¼   a2c T 4 þ   (23)
2 β þ αθ βη g1 α; β; η; θ

where α, β, η, and θ are components of the steady-state Kalman gain:

α Tη


K∞ ¼ (24)
β=T θ

and,
     
g1 α; β; η; θ ¼ βη αθ β αθ βη α θ 4 2α β 2θ þ αθ βη (25)
240 Kalman Filters - Theory for Advanced Applications

g2 α; β; η; θ ¼ α3 θðθ 1Þ þ α2 β 2 3ηθ2 þ 6ηθ 2η

   
2Þðθ 2θ
þ α2 θ2 ð2 2Þ þ αβ2 1 þ 2η θ 3η2 þ 3η2 θ
 
θÞ þ 2αβθðη 2Þðθ (26)
þ β3 ηð1 þ ηÞð1 ηÞ þ β 2 ηð 2 ηÞ ð θ 2Þ þ β2 ð2 θÞ

g3 α; β; η; θ ¼ αθ 2η2 þ 2ηθ þ θ2 2η2 θ2 2ηθ þ θ2 ð2

     
θ þ ηβ 2η þ 2θ θÞ (27)

Eq. (22) is obtained from σ2p and eac of the steady-state PVM Kalman filter (α β η θ filter)
by using Eq. (17). The derivation processes for these are shown in Ref. [25]. The relationship
between the steady-state Kalman gains and Qgen is derived as follows:

T 2 Bv  2
αβ þ 2β3 þ β2 þ ðθ 1Þθβ Rxv

a¼ 2
1 Rxv β ð1 θÞα θ

þα2 Rxv ð1 θÞ 1 þ 2αβ þ β2 θ þ β 3θ 2 θ2 þ θð1 θÞðα

    

1Þ (28)

β3 R2xv þ Rxv α βð1 θ2 þ θ þ β2 θ þ βθ þ θ2

   
θÞ θ
b¼ TBv (29)
1 Rxv β2 ð1 θÞα θ

Rxv αβθ þ β2 ðθ þ 1Þ αθ β þ θ þ βθ þ θ2
   
βθ
c¼ Bv (30)
1 Rxv β2 ð1 θÞα θ

where

Rxv ¼ Bx =T 2 Bv (31)

η ¼ Rxv β (32)

The derivation of these is given in the Appendix. Note that the dimensionless parameter Rxv
corresponds to the ratio of the measurement accuracies in position and velocity and directly
affects the tracking accuracy in PVM tracking systems. From these results, we also obtain the
closed form of the RMS index for PVM systems and can design optimal Q using Eqs. (22)–(32).

4. Filter design strategy based on the RMS index

Using the RMS index introduced in the previous section, we can design the Kalman filter
parameters (i.e., Q) to achieve optimal tracking. This section defines the optimization problems
for POM and PVM systems with a Q that minimizes the RMS index εp .

4.1. RMS-index minimization problem

4.1.1. POM system optimization
The evaluating function to determine optimal Q is εp, pom normalized by Bx, which is defined as
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 241
http://dx.doi.org/10.5772/intechopen.71731

aD 2 2α2 þ 2β þ αβ
μpom ¼ ε2p, pom =Bx ¼ þ  (33)
β2

α 4 2α β

where

a2D ¼ a2c T 4 =Bx (34)

is the preset parameter for the proposed strategy. Substituting Eqs. (20)–(22) into (33), we
obtain μpom ða; b; c; aD Þ. Using this, the optimal (a, b, c) for the POM system is determined by
solving

arg min μpom ða; b; c; aD Þ

a, b, c (35)
sub: to: a > 0, b > 0, c > 0, and aD ¼ Const:

4.1.2. PVM system optimization

Like the POM system, a normalized RMS index can be used for the design of the PVM system.
Normalizing Eq. (22) by Bx and substituting Eqs. (31) and (32) into this, the evaluating function
for the PVM system is given by
!2    
2 2βRxv θ g2 α; β; θ; Rxv þ g3 α; β; θ; Rxv =Rxv
μpvm ¼ ε2p, pvm =Bx ¼ aD 2
þ (36)
2 β þ αθ β2 Rxv
   
g1 α; β; θ; Rxv

To design optimal (a, b, c) for the PVM system, the optimal steady-state Kalman gains are
calculated by solving the following minimization problem.
 
arg min μpvm α; β; θ; Rxv
α, β , θ (37)
sub: to: Stability conditions are satisfied, and aD ¼ Const:

where the stability conditions with respect to Kalman gains are easily derived by the well-
known Jury’s test as
 
ð1 ηÞβ < αθ and 4 2α β 2θ þ αθ ηβ > 0 and αθ ηβ α θ þ 1 < 1 (38)

Substituting the optimal (α, β, θ) calculated by Eq. (37) into Eqs. (28)–(30), we obtain an optimal
(a, b, c) for the PVM Kalman filter.

4.2. Procedure and notes of the proposed design strategy

The procedure of the proposed strategy for each system is summarized in this section.

4.2.1. Design procedure for a POM system

1. Set Bx from the sensor performance.
242 Kalman Filters - Theory for Advanced Applications

2. Preset aD based on the approximate target acceleration.

3. Determine (a, b, c) by solving Eq. (35).
The methodology of presetting aD is discussed in the simulation section.

4.2.2. Design procedure for a PVM system

1. Set Bx and Bv from the sensor performance.
2. Preset aD based on the approximate target acceleration.


3. Determine α; β; θ by solving Eq. (37).


4. Determine (a, b, c) from α; β; θ using Eqs. (28)–(30).

4.2.3. Notes on computation in the proposed strategy

With respect to the proposed strategy, note that:

• Eqs. (35) and (37) can be solved by gradient descent with several initial values. This is
because that the parameter searching range is narrow due to the stability conditions.

• The proposed design process is only carried out once before using the Kalman filter.
Although the computational cost of the above optimization process is not small, it does
not affect the Kalman filtering process.

4.3. Discussion on preset parameter aD

Here, the appropriate presetting for aD in practical use is discussed. The covariance matrix of
process noise Q determined by the proposed strategy is only optimal when aD is matched to
the target acceleration and the target is moving with constant acceleration corresponding to aD .
However, using the proposed strategy, the tracking accuracy is always better than when using
conventional models as verified in Ref. [3]. Consequently, the proposed method achieves
sufficient accuracy, even if aD is not matched to the true target acceleration. This means that
the relatively small difference between the true and preset acceleration is acceptable. Thus, in
practical use, we estimate an approximate or a typical value for the acceleration (e.g., mean
and maximum) in advance based on the assumed motion of the target and then set aD by using
this estimated value. The example application presented in Section 6 assumes the approximate
maximum acceleration of the target is known and is used for the Kalman filter design.

Thus, target acceleration information is required for accurate Kalman filter tracking by using
the proposed strategy. As a method to obtain an approximated acceleration, communications
between the tracking systems and the accelerometers embedded in targets can be considered.
Many sensing targets have acceleration sensors; for example, robots and vehicles have inertial
sensors, and humans have accelerometers embedded in smartphones. Soon, Internet of Things
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 243
http://dx.doi.org/10.5772/intechopen.71731

technology will make data communications between robots, smartphones, and radar possible.
Thus, we can obtain approximated acceleration based on this novel technology.

5. Theoretical steady-state performance analysis

This section presents theoretical performance analyses of the Kalman tracking filters by using
the proposed design strategy. With respect to POM systems, our previous study [3] verified the
effectiveness of the proposed strategy by comparison with a conventional random acceleration
model based filter design. Thus, the RMS indices for the following filters are compared:
• Optimal POM filter: the Kalman filter for the POM system designed using the strategy
mentioned in Section 4.2.1.
• Optimal PVM filter: the Kalman filter for the PVM system designed using the strategy
mentioned in Section 4.2.2.
• RA filter: the Kalman filter for the PVM system with the RA process noises by using
optimal σq with respect to the RMS index.

The comparison of the optimal PVM filter with the RA filter indicates the effectiveness of the
proposed strategy (i.e., considering the arbitrary covariance matrix of the process noise Qgen )
and the comparison of the optimal POM and PVM filters illustrates the enhancement of
tracking accuracy by using the velocity measurements in the proposed strategy. This section
assumes that Bx and T are normalized to 1.
Figure 1 shows the relationship between the design parameter aD and the minimum RMS
index εp, opt for Rxv = 1 (Figure 1 left) and Rxv = 10 (Figure 1 right). It can be seen that the
optimal PVM filter achieves the best performance. This result verifies that the proposed strat-
egy determines steady-state gains corresponding to a better covariance matrix of process noise
than the RA model. The optimal PVM filter also achieves better performance compared with
the optimal POM filter even for Rxv ¼ 1, which means that the measurement accuracy of the

Figure 1. Analytical relationship between aD and εp, opt (Rxv ¼ 1 (left), Rxv ¼ 10 (right)).
244 Kalman Filters - Theory for Advanced Applications

Figure 2. Analytical relationship between Rxv and εp, opt (aD ¼ 0:01 (left), aD ¼ 0:1 (right)).

position and velocity is the same. The addition of the velocity measurements effectively
enhances the tracking accuracy. Furthermore, when the velocity measurement accuracy is high,
the optimal PVM filter achieves greater accuracy than the POM filter.

Figure 2 shows the relationship between Rxv and εp, opt for a2D ¼ 0:01 (left) and 0.1 (right). Both
cases exhibit the same trend. For both optimal PVM and RA filters, better performance is
achieved with better velocity measurement accuracy. The performance of the optimal PVM
filter is better than that of the optimal POM filter including relatively small Rxv (the velocity
measurement accuracy is low). In contrast, the performance of the RA filter is worse than that
of the optimal POM filter for small Rxv because the covariance matrix is limited to Eq. (14).
Moreover, by comparing the two insets of Figure 2, we see the greater effectiveness of the
proposed strategy for relatively large aD .

6. Application to radar tracking simulation

Finally, this section provides an example of the Kalman filter tracker designed with the
proposed strategy in a realistic application, namely, pulse Doppler radar tracking.

6.1. Simulation setup

We simulated the pulse Doppler radar tracking of a maneuvering target and compared the
tracking errors of the filters assumed in the previous section. Figure 3 shows the simulation
 
scenario and the true target acceleration. The true target position is xtk ; ytk ¼ ðkT Þ2 ; 20þ


ðkT Þ1:5 cos ðπkT=5ÞÞ. Two-dimensional tracking in the x-y plane of the point target is assumed.
We consider two pulse Doppler radars located at (x, y) = (0.5 m, 0) and (1.0 m, 0). The sampling
interval T is 100 ms, and the observation time is 4 s. The transmitted signal is a pulse with
central frequency of 60 GHz and bandwidth of 500 MHz. The received radar signals are
calculated using ray tracing with the addition of the Gaussian white noise. The radar measure-
ment parameter depends on the system under consideration: the POM system assumes the
measurement of the position by using ranging results, and the PVM system assumes the
position and velocity measurements where the position measurement is the same as the POM
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 245
http://dx.doi.org/10.5772/intechopen.71731

Figure 3. Simulation setting (simulation scenario (left), true target acceleration (right)).

system, and the velocity measurement is based on the Doppler shift with the method
presented in Ref. [18]. We determine a variance for this noise to set Bx ¼ 9  10 4 m2 and
Bv ¼ 0:09 m2/s2. In these settings, Rxv ¼ 1. These values are the averages along the two axes.
Using the RMS prediction error calculated from 1000 Monte Carlo simulations, the perfor-
mance is defined as
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1000 
u 2
u 1 X 2 
εk ¼ t xtk xpmk þ ytk ypmk (39)
1000 m¼1

where xpmk and ypmk are the predicted positions in the mth Monte Carlo simulation.

6.2. Implementation of Kalman filter

First, the implementation of the Kalman filters for two-dimensional system is presented. The
implementation of a two-dimensional optimal POM filter is as follows:
 T
xt ¼ xt vxt yt vyt (40)

1 T 0 0
0 1

B0 1 0 0C
B C
Φ¼B
B0 0
C (41)
@ 1 TC
A
0 0 0 1
!
1 0 0 0
H¼ (42)
0 0 1 0

Bx 0


R¼ (43)
0 Bx
246 Kalman Filters - Theory for Advanced Applications

aopt bopt 0 0
0 1
B bopt copt 0 0 C
Q¼B (44)
B C
@0 0 aopt bopt A
C

0 0 bopt copt

where vyt is the true velocity in the y-axis and aopt ; bopt ; copt is optimized (a, b, c), calculated
using the procedure in Section 4.2.1. xt and Φ of a two-dimensional PVM filter are the same as
for a POM filter. H and R are

1 0 0 0
0 1
B0 1 0 0C
H¼B (45)
B C
C
@0 0 1 0A
0 0 0 1

Bx 0 0 0
0 1
B0 Bv 0 0 C
R¼B (46)
B C
C
@0 0 Bx 0 A
0 0 0 Bv


In addition, the formulation of Q is the same as in Eq. (43) and aopt ; bopt ; copt is calculated
using the procedure in Section 4.2.2. A two-dimensional RA filter is the same as the optimal
PVM filter, with the exception of Q. Q of the RA filter is

T 4 =4 T 3 =2 0
0 1
0
B T 3 =2 T 2 0 0 C
C 2
Qra ¼ B Cσq (47)
B
4 3
@0 0 T =4 T =2 A
0 0 T 3 =2 T 2

Next, the design for an appropriate aD is presented. We presume an approximate prediction of

accelerations. For instance, when the maximum acceleration of the target in Figure 3 is
predicted to be approximately ac ¼ 3 m/s2, aD is then 1.0, from Eq. (34). Using this aD and the


radar settings described in the previous section, we have aopt ; bopt ; copt for each filter.

6.3. Results and discussion

Figure 4 shows the simulation results. Clearly, the filters using velocity measurements achieve
greater accuracy than the optimal POM filter. The mean steady-state prediction RMS errors
(E½εk Š in 2 s < kT) of the optimal POM, RA, and optimal PVM filters are 0.59, 0.46, and 0.19 m,
respectively. These results indicate that the proposed strategy achieves greater accuracy than
the conventional RA filter even in realistic situations. The mean RMS error of the optimal PVM
filter is 41% of that of the RA filter. This is because the RA model cannot track the abrupt
motion of the high-maneuvering target because of limitations in expressing the process noise.
In contrast, the optimal PVM filter can set gains corresponding to the appropriate process
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 247
http://dx.doi.org/10.5772/intechopen.71731

Figure 4. Simulation results.

noise to accurately track high-maneuvering target. Moreover, the mean RMS error of the
optimal PVM filter is 32% of the error in optimal POM filter, and this clearly indicates the
effectiveness of the velocity measurement, even when Rxv ¼ 1 (when the measurement reli-
ability of the position and velocity are the same). These simulation results are consistent with
the theoretical analyses presented in Figure 1.

7. Final remarks

7.1. Conclusions
In this chapter, the efficient steady-state performance index, known as the RMS index, was
introduced for both POM and PVM Kalman filters for systems that track moving objects.
Automatic design (preset) of the covariance matrix of the process noise Q, to realize optimal
position prediction, was achieved using the analytical relationship between Q and the RMS
index. The validity of the proposed design strategy was shown via analyses and simulations.
These results verified that the proposed index attained accurate tracking when compared with
the conventional RA-model-based Kalman filter design. A simulation of a realistic situation
indicated that the optimal performance given by the proposed strategy is 41% better than that
given by the conventional design procedure for a PVM system. Moreover, the optimal perfor-
mance of the optimal POM and PVM Kalman filters was compared showing that the optimal
PVM Kalman filter is accurate when compared with the POM filter in a steady state.

7.2. Future works

The most important future objective is the extension of the RMS index-based design strategy
to the third-order (and higher order) Kalman filters that are widely used for real applications.
In third-order tracking, an acceleration is added to the state vector, becoming one of the
input parameters of the Kalman filter. Performance analysis and the establishment of a design
248 Kalman Filters - Theory for Advanced Applications

strategy for such systems (i.e., position/acceleration and position/velocity/acceleration mea-

sured Kalman filters) are important considerations for advanced sensor fusion systems
under development. Moreover, considerations of other dynamic models (e.g., the constant
turn model) should also be probed for use in many applications including pedestrian tracking.

A. Appendix

A.1. Derivation of Eqs. (28)–(30)

Because we assume a steady state, the index k of all parameters and matrices is omitted in the
following calculations. The ith row and jth column of a matrix P are denoted as Pi, j .
b as
Eq. (11) is also written using P

b TR
K ¼ PH 1
(48)

As indicated in Eq. (7), H of the PVM Kalman filter is the identity matrix. Thus, from Eq. (48),
the relationship between the Kalman gains and the error covariance matrix in the estimated
b is calculated using Eqs. (8) and (24) as
state P
 
αBx TηBv
b ¼ KR ¼
P (49)
βBx =T θBv

With P1, 2 ¼ P2, 1 and Eq. (31), we have the following relationship:

Bx
η¼β ¼ βRxv (50)
T 2 Bv

Eq. (50) is equal to Eq. (32), showing that this relationship is satisfied in the assumed PVM
Kalman filter without depending on the process noise. P b is also calculated using Eq. (13) by
substituting Eqs. (7) and (24) as
!
ð1 ~ 1, 1
αÞP ~ 1, 2
TηP ð1 ~ 1, 2
α ÞP ~ 2, 2
TηP
b¼
P
1, 1
1, 2 (51)
ð1 ~ 1, 2
θÞP β=T P ~ ð1 ~ 2, 2
θ ÞP β=T P ~

~
Elements of Pare required to calculate Eq. (51) and are calculated using Eqs. (3), (12), and (18) as
!
1, 1 1, 2 2 2, 2 1, 2 2, 2
~¼ P þ 2TP þ T P þ a P þ TP þ b
P (52)
P1, 2 þ TP2, 2 þ b P2, 2 þ c

Substituting Eq. (52) into Eq. (51), and comparing elements of Eq. (49), we have the following
linear system:
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 249
http://dx.doi.org/10.5772/intechopen.71731

αÞ αBx þ 2ηT 2 Bv þ θT 2 Bv þ a


αBx ¼ ð1 ηT ðηTBv þ θTBv þ bÞ (53)

βBx =T ¼ ð1 αÞðηTBv þ θTBv þ bÞ ηT ðθBv þ cÞ (54)

θBv ¼ ð1 θÞðθBv þ cÞ β=T ðηTBv þ θTBv þ bÞ (55)

Solving this linear system with respect to (a, b, c) and substituting Eq. (50) into the solutions,
we arrive at Eqs. (28)–(30).

Author details

Kenshi Saho
Address all correspondence to: [email protected]
Department of Intelligent Systems Design Engineering, Toyama Prefectural University, Imizu,
Toyama, Japan

References

[1] Ekstrand B. Some aspects on filter design for target tracking. Journal of Control Science
and Engineering. 2012;2012:870890. DOI: 10.1155/2012/870890
[2] Bar-Shalom Y, Li XR, Kirubarajan T. Estimation and Tracking: Principles, Techniques, and
Software. Boston, MA, United States: Artech House Publishers; 1998
[3] Saho K, Masugi M. Automatic parameter setting method for an accurate Kalman filter
tracker using an analytical steady-state performance index. IEEE Access. 2015;3:1919-1930.
DOI: 10.1109/ACCESS.2015.2486766
[4] Hashlamon I, Erbatur K. An improved real-time adaptive Kalman filter with recursive
noise covariance updating rules. Turkish Journal of Electrical Engineering & Computer
Sciences. 2016;24:524-540. DOI: 10.3906/elk-1309-60

[5] Basso G, Amorim TG, Brito AV, Nascimento TP. Kalman filter with dynamical setting of
optimal process noise. IEEE Access. 2017;5:8385-8393. DOI: 10.1109/ACCESS.2017.2697072

[6] Fan Y, Lu F, Zhu W, Bai G, Yan L. A hybrid model algorithm for hypersonic glide vehicle
Maneuver tracking based on the aerodynamic model. Applied Sciences. 2017;7:159. DOI:
10.3390/app7020159

[7] Li W, Sun S, Jia Y, Du J. Robust unscented Kalman filter with adaptation of process and
measurement noise covariances. Digital Signal Processing. 2016;48:93-103. DOI: 10.1016/j.
dsp.2015.09.004
250 Kalman Filters - Theory for Advanced Applications

[8] Vivone F, Braca P, Horstmann J. Knowledge-based multitarget ship tracking for HF surface
wave radar systems. IEEE Transactions on Geoscience and Remote Sensing. 2015;53:3931-
3949. DOI: 10.1109/TGRS.2014.2388355
[9] Du G, Zhang P. A Markerless human robot Interface using particle filter and Kalman
filter for dual robots. IEEE Transactions on Industrial Electronics. 2015;62:2257-2264. DOI:
10.1109/TIE.2014.2362095
[10] Martino L, Read J, Elvira V, Louzada F. Cooperative parallel particle filters for on-line
model selection and applications to urban mobility. Digital Signal Processing. 2017;60:
172-185. DOI: 10.1016/j.dsp.2016.09.011
[11] Drovandi CC, McGree J, Pettitt AN. A sequential Monte Carlo algorithm to incorporate
model uncertainty in Bayesian sequential design. Journal of Computational and Graphi-
cal Statistics. 2014;23:324. DOI: 10.1080/10618600.2012.730083
[12] Li XR, Jilkov VP. Survey of maneuvering target tracking. Part I: Dynamic models. IEEE
Transactions on Aerospace and Electronic Systems. 2003;39:1333-1364 DOI: 10.1109/
TAES.2003.1261132
[13] Crouse DF. A general solution to optimal fixed-gain (alpha-beta-gamma etc.) filters. IEEE
Signal Processing Letters. 2015;22:901-904. DOI: 10.1109/LSP.2014.2376876

[14] Tang X, Falco G, Falletti E, Presti LL. Complexity reduction of the Kalman filter-based
tracking loops in GNSS receivers. GPS Solutions. 2017;21:685-699. DOI: 10.1007/s10291-
016-0557-6

[15] Jatoth RK, Gopisetty S, Hussai M. Performance analysis of alpha beta filter, Kalman filter
and meanshift for object tracking in video sequences. International Journal of Image,
Graphics and Signal Processing. 2015;3:24-30. DOI: 10.5815/ijigsp.2015.03.04

[16] Li Z, Wu H. A survey of Maneuvering target tracking using Kalman filter. Proc. 4th Int.
Conf. Mechatronics Mater. Chem. Comput. Eng. (ICMMCCE). 2015;39:542-545. DOI:
10.2991/icmmcce-15.2015.109

[17] Saho K, Homma H, Sakamoto T, Sato T, Inoue K, Fukuda T. Accurate image separation
method for two closely spaced pedestrians using UWB Doppler imaging radar and
supervised learning. IEICE Transactions on Communications. 2014;97:1223-1233. DOI:
10.1587/transcom.E97.B.1223

[18] Anabuki M, Okumura S, Sato T, Sakamoto T, Saho K, Yoshioka M, Inoue K, Fukuda T,

Sakai H. Ultrawideband radar imaging using adaptive Array and Doppler separation.
IEEE Transactions on Aerospace and Electronic Systems. 2017;53:190-200. DOI: 10.1109/
TAES.2017.2649798

[19] Cossaboom M, Georgy J, Karamat T, Noureldin A. Augmented kalman filter and map
matching for 3D RISS/GPS integration for land vehicles. International Journal of Naviga-
tion and Observation. 2012;2012:576807. DOI: 10.1155/2012/576807
 Kalman Filter for Moving Object Tracking: Performance Analysis and Filter Design 251
http://dx.doi.org/10.5772/intechopen.71731

[20] Afia AB, Escher AC, Macabiau C, Roche S. A GNSS/IMU/WSS/VSLAM hybridization

using an extended Kalman filter. In: Proc. of ION 2015 Pacific PNT Meeting. 2015.
pp. 719-732
[21] Kellner D, Barjenbruch M, Klappstein J, Dickmann J, Dietmayer K. Tracking of extended
objects with high-resolution Doppler radar. IEEE Transaction on Intelligent Transporta-
tion Systems. 2016;17:1341-1353. DOI: 10.1109/TITS.2015.2501759
[22] Carrolla P, Domresea K, Zhoub H, Zhoua S, Willetta P. Doppler-aided localization of
mobile nodes in an underwater distributed antenna system. Physical Communication.
2016;18:49-59. DOI: 10.1016/j.phycom.2015.08.008
[23] Zhoua G, Wub L, Xiea J, Denga W, Quan T. Constant turn model for statically fused
converted measurement Kalman filters. Signal Processing. 2015;108:400-411. DOI: 10.1016/
j.sigpro.2014.09.022
[24] Cho MH, Tahk MJ, Kim MC, Lee D, Yoon J. Modified gain pseudo-measurement filter
design for radar target tracking with range rate measurement. In: Proc. IEEE 2017 25th
Mediterranean Conf. on Control and Automation (MED). 2017. pp. 1195-1200. DOI: 10.1109/
MED.2017.7984280
[25] Saho K, Masugi M. Performance analysis of fixed-gain position-velocity-measured track-
ing filter. In: 2016 2nd International Conference on Control, Automation and Robotics
(ICCAR); 28-30 April 2016; Hong Kong, China. IEEE; 2016. DOI: 10.1109/ICCAR.2016.
7486760
[26] Saho K, Masugi M. Performance analysis of alpha-beta-gamma tracking filters using
position and velocity measurements. EURASIP Journal on Advances in Signal Processing.
2015;2015:35. DOI: 10.1186/s13634-015-0220-3
[27] Kalata PR. The tracking index: A generalized parameter for alpha-beta and alpha-beta-
gamma target trackers. IEEE Transactions on Aerospace and Electronic Systems. 1984;
AES-20:174-182. DOI: 10.1109/TAES.1984.310438