Professional Documents
Culture Documents
Full Ebook of Synergies in Analysis Discrete Mathematics Soft Computing and Modelling 1St Edition P V Subrahmanyam Online PDF All Chapter
Full Ebook of Synergies in Analysis Discrete Mathematics Soft Computing and Modelling 1St Edition P V Subrahmanyam Online PDF All Chapter
Full Ebook of Synergies in Analysis Discrete Mathematics Soft Computing and Modelling 1St Edition P V Subrahmanyam Online PDF All Chapter
https://ebookmeta.com/product/probabilistic-methods-in-discrete-
mathematics-v-f-kolchin-editor/
https://ebookmeta.com/product/progress-in-pure-and-applied-
discrete-mathematics-vol-1-probabilistic-methods-in-discrete-
mathematics-2nd-edition-v-f-kolchin/
https://ebookmeta.com/product/soft-computing-and-signal-
processing-proceedings-of-3rd-icscsp-2020-volume-2-1340-advances-
in-intelligent-systems-and-computing-1340-v-sivakumar-reddy-
editor/
https://ebookmeta.com/product/discrete-mathematics-inc-barcharts/
Advanced Soft Computing Techniques in Data Science, IoT
and Cloud Computing Sujata Dash (Editor)
https://ebookmeta.com/product/advanced-soft-computing-techniques-
in-data-science-iot-and-cloud-computing-sujata-dash-editor/
https://ebookmeta.com/product/discrete-mathematics-essentials-
and-applications-1st-edition-ali-grami/
https://ebookmeta.com/product/soft-computing-techniques-in-
engineering-health-mathematical-and-social-sciences-edge-ai-in-
future-computing-1st-edition-pradip-debnath/
https://ebookmeta.com/product/geometry-and-discrete-
mathematics-2nd-edition-benjamin-fine/
https://ebookmeta.com/product/discrete-mathematics-3rd-edition-
oscar-levin/
Forum for Interdisciplinary Mathematics
P. V. Subrahmanyam
V. Antony Vijesh
Balasubramaniam Jayaram
Prakash Veeraraghavan Editors
Synergies in
Analysis, Discrete
Mathematics,
Soft Computing
and Modelling
Forum for Interdisciplinary Mathematics
Editors-in-Chief
Viswanath Ramakrishna, University of Texas, Richardson, USA
Zhonghai Ding, University of Nevada, Las Vegas, USA
Editorial Board
Ashis SenGupta, Indian Statistical Institute, Kolkata, India
Balasubramaniam Jayaram, Indian Institute of Technology, Hyderabad, India
P. V. Subrahmanyam, Indian Institute of Technology Madras, Chennai, India
Ravindra B. Bapat, Indian Statistical Institute, New Delhi, India
The Forum for Interdisciplinary Mathematics is a Scopus-indexed book series. It
publishes high-quality textbooks, monographs, contributed volumes and lecture
notes in mathematics and interdisciplinary areas where mathematics plays a
fundamental role, such as statistics, operations research, computer science, financial
mathematics, industrial mathematics, and bio-mathematics. It reflects the increasing
demand of researchers working at the interface between mathematics and other
scientific disciplines.
P. V. Subrahmanyam · V. Antony Vijesh ·
Balasubramaniam Jayaram ·
Prakash Veeraraghavan
Editors
Synergies in Analysis,
Discrete Mathematics, Soft
Computing and Modelling
Editors
P. V. Subrahmanyam V. Antony Vijesh
Department of Mathematics Department of Mathematics
Indian Institute of Technology Madras Indian Institute of Technology Indore
Chennai, India Indore, Madhya Pradesh, India
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
Mathematics permeates every aspect of our daily life, and it can aptly be called
“God’s own science”. The Forum for Interdisciplinary Mathematics, Stella Maris
College, and S.S.N College of Engineering jointly organized an online international
conference entitled “Synergies in Computational, Mathematical, Statistical and Phys-
ical Sciences (FIM2020: SCMPS2020), from 23–27 November 2020 at Chennai,
India, highlighting the interdisciplinary nature of mathematics. This conference
attracted about 250 researchers, both eminent and emergent, and comprised numerous
keynote addresses, plenary talks, symposia on special areas and parallel sessions
of paper presentation. For the convenience of potential authors, several venues for
refereed publication of submitted articles were suggested by the conference orga-
nizing committee, including the present proceedings in the Springer’s Forum for
Interdisciplinary Mathematics series.
Both the invited articles and submitted papers were broadly grouped under three
parts: analysis and modelling, discrete mathematics and applications, and fuzzy set
theory and applications. More than fifty papers were submitted and more than 100
referees of international standing evaluated the submissions critically despite the
constraints due to the prevalent pandemic. Each paper was at least doubly refereed
with the acceptance rate being one out of three. The authors of those papers which
could not be accepted were given useful suggestions for improvement which will
hopefully help them publish the revised versions elsewhere.
We place on record our thanks to the referees for sparing their valuable time
and energy in critically appraising the articles submitted. We also acknowledge the
support provided by the management of the Forum for Interdisciplinary Mathematics
series in publishing the present proceedings.
v
Contents
vii
viii Contents
ix
x About the Editors
Let H denote the class of functions that are analytic in U = {z ∈ C : |z| < 1}. If
f ∈ A, then f is analytic in U, and
f (z) = z + a2 z 2 + a3 z 3 + · · · . (1.1)
P. Gurusamy
Velammal Engineering College, Surapet, Chennai 600066, Tamil Nadu, India
R. Jayasankar
Government College of Arts and Science, Arakkonam, Tamil Nadu, India
S. Sivasubramanian (B)
Department of Mathematics, University College of Engineering Tindivanam, Anna University,
Tindivanam 604001, India
e-mail: sivasaisastha@rediffmail.com
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 1
P. V. Subrahmanyam et al. (eds.), Synergies in Analysis, Discrete Mathematics, Soft
Computing and Modelling, Forum for Interdisciplinary Mathematics,
https://doi.org/10.1007/978-981-19-7014-6_1
2 P. Gurusamy et al.
and
z f (z)
C(β) = h ∈ A : 1 + > β, 0 ≤ β < 1, z ∈ U
f (z)
denote the renowned classes of starlike functions of order β, and convex functions
of order β. Make a note of that S∗ = S∗ (0) is called the class of starlike func-
tions (with respect to the origin). The classes S∗ (β) and C(β) were introduced and
discussed in detail by Robertson [18]. It is to be prominent that in the terminol-
z f (z) 1+z
ogy of subordination, the class S∗ = S∗ (0) can be denoted by ≺ .
f (z) 1−z
∗ ∗
Mendiratta et al. [15] considered the class of starlike functions Se = S (e ) defined
z
z f (z)
by ≺ e z z ∈ U. In [7], the Hermitian Toeplitz matrix Tq,n ( f ), q, n ∈ N of
f (z)
a function f ∈ A of the form (1.1) is defined by
⎡ ⎤
an an+1 ... an+q−1
⎢ a n+1 an ... an+q−2 ⎥
⎢ ⎥
⎢
Tq,n ( f ) = ⎢ . . ... . ⎥ ⎥
⎣ a n+q−1 a n+q−2 ... an ⎦
,
and obtained the estimates of T2 [n], T3 [1], T3 [2] and T2 [3] over few subclasses
of A. In recent times, many articles have been published in finding bounds of
determinants, whose elements are coefficients of functions in f ∈ A or its sub-
classes, to name a few Hankel matrices, i.e., square matrices which have con-
stant entries along the reverse diagonal and the generalized Zalcman functional
Jm,n ( f ) = am+n−1 − am an , m, n ∈ N, are of particular interest. From the collec-
tion of the number of articles available in this direction, one may refer to [2–6, 9–14,
17, 19] where the authors have discussed the bounds of second- and third-order
Hankel determinants. Cudna et al. [7] established some elementary properties of the
Hermitian Toeplitz determinant Tq,1 ( f ) .
Given a subclass G of A, let A2 (G) = max {|a2 | : f ∈ G}. Thus, for f ∈ A,
T2,1 ( f ) = 1 − |a2 |2 ,
1 The Second- and Third-Order Hermitian Toeplitz Determinants … 3
and therefore, the results are mentioned by a function in G which is extremal and for
the upper bound when f is the identity function.
In the present article, we will get the estimates T2,1 ( f ) and T3,1 ( f ) , for few
new subclasses S∗s (e z ), Cs (e z ), S∗ (e z ) and C (e z ),
where T2,1 ( f ) = 1 − |a2 |2 and
1 a2 a3
T3,1 ( f ) = a2 1 a2 = 2 a22 a 3 − 2|a2 |2 − |a3 |2 + 1. (1.2)
a3 a2 2
To prove the main theorems, we need the following Lemmas which we state now.
Let P denote the class of functions p analytic in U for which { p (z)} > 0,
∞
p (z) = 1 + pn z n , z ∈ U. (1.3)
n=1
for some ζ ∈ U
Lemma 1.2 ([7]) Let G be a subclass of the class A, where A2 (G) exists. If the
identity is an element of G, then
1 + ω (z)
p (z) = = 1 + p1 z + p2 z 2 + . . . ,
1 − ω (z)
analytic in U with p(0) = 1 and maps U onto the right half of the ω-plane.
Computing ω (z) in terms of p(z), we get
p (z) − 1
ω (z) = .
p (z) + 1
Now,
(ω (z))2 (ω (z))3
eω(z) = 1 + ω (z) + + + ··· . (1.7)
2 6
From (1.6) and (1.7), we have
ω(z) p1 z p2 p2 p3 p1 p2 p3
e =1+ + − 1 z +
2
− + 1 z3 + · · · . (1.8)
2 2 8 2 4 48
z f (z)
≺ ez z ∈ U. (1.1)
f (z)
1
a2 = p1 (1.3)
2
and
1 1 2
a3 = p2 + p . (1.4)
4 16 1
Since
A2 S∗ (e z ) = 1,
0 ≤ T2,1 ( f ) ≤ 1. (1.5)
Both inequalities are sharp. Let e be a holomorphic function which is the solution
of the differential equation:
1 The Second- and Third-Order Hermitian Toeplitz Determinants … 5
ς e (ς )
= eς , ς ∈ D, e (0) = 0, e (0) = 1,
e (ς )
i.e.,
ς
eζ − 1 3 17
e (ς ) = ς exp dz = z + z 2 + z 3 + z 4 + · · · , z ∈ D. (1.6)
0 z 4 36
1
− ≤ T3,1 ( f ) ≤ 1. (1.7)
16
Both inequalities are sharp. The result is sharp for the function defined in (1.6).
Proof Since the class f ∈ S∗ (e z ) and T3,1 ( f ) are rotationally invariantly, we may
assume that p = p1 ∈ [0, 2]. Thus, we infer from (1.2), (1.3) and (1.4), that
1 2 7 4 1 3 2
T3,1 ( f ) = 1 − p + p − | p2 |2 + p ( p2 ) .
2 256 16 32
Using (1.4), we have
2
1 2 15 4 4 − p 2 |ζ |2 4 − p 2 p 2 |ζ | 3 4 − p 2 p 2 (ζ )
T3,1 ( f ) = 1 − p + p − − + .
2 256 64 32 64
(1.8)
= ψ ( p, |ζ | , (ζ )) ζ ∈ U.
x 15x 2 (4 − x)2 y 2 (4 − x) x y
G (x, y) = 1 − + − + . (1.10)
2 256 64 64
Case I(i): For x = 0 in (1.10), we obtain
y2
G (0, y) = 1 − ≤ 1, y ∈ [0, 1] .
4
6 P. Gurusamy et al.
1
G (4, y) = − < 1, y ∈ [0, 1] .
16
Case I (iii): For y = 0 in (1.10), we obtain
x 15x 2
G (x, 0) = 1 − + = γ (x) , x ∈ [0, 4] .
2 256
Since
1
γ (0) = − <0
2
and
1
γ (4) = − <0
32
So γ is decreasing function on [0,4]. Therefore,
3 5x 7x 2
G (x, 1) = − + = ϕ (x) x ∈ [0, 4] .
4 16 256
Since
5
ϕ (0) = − <0
16
and
24
γ (4) = − <0
256
So, ϕ is decreasing function on [0,4]. Therefore,
3
G (x, 1) = ϕ (x) ≤ ϕ (0) = < 1, x ∈ [0, 4] .
4
Case I(v): Let (x, y) ∈ (0, 4) × (0, 1). Then
∂G (4 − x)2 2y (4 − x) x
=− + =0
∂y 64 64
if and only if
x
y= .
2 (x − 4)
1 The Second- and Third-Order Hermitian Toeplitz Determinants … 7
the solution does not belong to (0,4). Therefore G has no critical points in (0, 4) ×
(0, 1).
Hence, from cases from V(i) to V(v), it follows that
3 5x 7x 2
ψ (x) = − + x ∈ [0, 4] .
4 16 256
Since
5
ψ (0) = − <0
16
and
3
ψ (4) = − < 0.
32
So, ψ is decreasing function on [0,4]. Therefore,
1
ψ (x) ≥ ψ (4) = − , x ∈ [0, 4] . (1.12)
16
This completes the proof of Theorem 1.2.
Definition 1.2 A function f (z), given by (1.1), is said to be in the class C(e z ) if the
following condition is satisfied:
z f (z)
1+ ≺ ez z ∈ U. (1.13)
f (z)
Since
1
A2 C(e z ) = .
2
Using the Lemma1.2, we have the following theorem.
8 P. Gurusamy et al.
3
≤ T2,1 ( f ) ≤ 1.
4
Both inequalities are sharp.
Now we find the bounds for T3,1 ( f ) .
Theorem 1.4 If f ∈ C (e z ), then
9
≤ T3,1 ( f ) ≤ 1. (1.14)
16
Both inequalities are sharp.
Proof Let f ∈ C (e z ), then it follows from (1.13) and using the principle of subor-
dination that
z f (z)
1+ = eω(z) (1.15)
f (z)
As explained earlier, as the class C(e z )and T3,1 ( f ) are rotationally invariantly,
we may assume that
p = p1 ∈ [0, 2]. Thus from (1.2), (1.16) and (1.17), gives
1 2 5 1 1 2
T3,1 ( f ) = 1 − p + p4 − | p 2 |2 + p Re ( p2 ) .
8 2304 144 144
By virtue of (1.4), we have
2
p2 p4 4 − p 2 |ζ |2 p 2 4 − p 2 |ζ | p 2 4 − p 2 (ζ )
T3,1 ( f ) = 1 − + − − + . (1.18)
8 256 576 288 288
= ψ ( p, |ζ | , (ζ )) ζ ∈ U.
x x2 (4 − x)2 y 2 (4 − x) x y
G (x, y) = 1 − + − − . (1.20)
8 256 576 144
Case III(i): Forx = 0 in (1.20), we obtain
y2
G (0, y) = 1 − ≤ 1, y ∈ [0, 1] .
36
Case III(ii): Forx = 4 in (1.20), we obtain
9
G (4, y) = < 1, y ∈ [0, 1] .
16
Case III(iii): Fory = 0 in (1.20), we obtain
x x2
G (x, 0) = 1 − + = γ (x) , x ∈ [0, 4] .
8 256
Since
1
γ (0) = − <0
8
and
3
γ (4) = − <0
32
So, γ is decreasing function on [0,4]. Therefore,
35 5x 21x 2
G (x, 1) = − + = ϕ (x) x ∈ [0, 4] .
36 36 2304
Since
5
ϕ (0) = − <0
36
and
19
ϕ (4) = − <0
288
So, ϕ is decreasing function on [0,4]. Therefore,
10 P. Gurusamy et al.
35
G (x, 1) = ϕ (x) ≤ ϕ (0) = < 1, x ∈ [0, 4] .
36
Case III(v): Let (x, y) ∈ (0, 4) × (0, 1). Then
∂G (4 − x)2 2y (4 − x) x
=− − =0
∂y 576 144
if and only if
2x
y= .
(x − 4)
the solution is not in (0, 4). Thus, G has no critical points in (0, 4) × (0, 1).
Hence, from cases from VII(i) to VII(v), it follows that
35 x 5x 2
ψ (x) = − + x ∈ [0, 4] .
36 9 2304
Since
1
ψ (0) = − <0
9
and
3
ψ (4) = − < 0.
32
So, ψ is decreasing function on [0, 4]. Therefore,
9
ψ (x) ≥ ψ (4) = , x ∈ [0, 4] . (1.22)
16
Let e be a holomorphic function which is the solution of the differential equation
1 The Second- and Third-Order Hermitian Toeplitz Determinants … 11
ς e (ς )
1+ = eς , ς ∈ D, e (0) = 0, e (0) = 1. (1.23)
e (ς )
This function e (ς ) satisfying (1.23) acts as the extremal function for the class
C∗ (e z ). This completes the proof of Theorem 1.4.
Definition 1.3 A function f (z), given by (1.1), is said to be in the class S∗s (e z ) if
the following condition is satisfied:
2 z f (z)
≺ ez z ∈ U. (1.24)
f (z) − f (−z)
If f ∈ S∗s (e z ), then it follows from (1.24) and using the principle of subordination
that
2 z f (z)
= eω(z) (1.25)
f (z) − f (−z)
From (1.6) and (1.7) and (1.25), (1.8) by equating corresponding coefficients, we
obtain
1
a 2 = p1 (1.26)
4
and
1 1 2
a3 = p2 − p . (1.27)
4 16 1
Since
1
A2 S∗s e z =
2
Using the Lemma 1.2, we have the following theorem.
Theorem 1.5 If f ∈ S∗s (e z ), then
3
≤ T2,1 ( f ) ≤ 1. (1.28)
4
Both inequalities are sharp.
Now we find the bounds for T3,1 ( f ) .
Theorem 1.6 If f ∈ S∗s (e z ), then
9
≤ T3,1 ( f ) ≤ 1. (1.29)
16
Both inequalities are sharp.
Proof Since the class f ∈ S∗s (e z ) and T3,1 ( f ) are rotated invariantly, we may
assume that p = p1 ∈ [0, 2]. Thus, from (1.2), (1.26), and (1.27), gives
12 P. Gurusamy et al.
1 2 3 4 1 1 2
T3,1 ( f ) = 1 − p − p − | p 2 |2 + p Re ( p2 ) .
8 256 16 16
Using (1.4), we have
2
1 2 1 4 4 − p 2 |ζ |2 p 2 4 − p 2 |ζ | p 2 4 − p 2 (ζ )
T3,1 ( f ) = 1 − p +
8 256
p −
64
+
32
+
32
. (1.30)
= ψ ( p, |ζ | , (ζ )) ζ ∈ U.
x x2 (4 − x)2 y 2
G (x, y) = 1 − + − . (1.32)
8 256 64
Case V (i): For x = 0 in (1.32), we obtain
y2
G (0, y) = 1 − ≤ 1, y ∈ [0, 1] .
4
Case V (ii): For x = 4 in (1.32), we obtain
9
G (4, y) = < 1, y ∈ [0, 1] .
16
Case V (iii): For y = 0 in (1.32), we obtain
x x2
G (x, 0) = 1 − + = γ (x) , x ∈ [0, 4] .
8 256
Since
1
γ (0) = − <0
8
and
3
γ (4) = − <0
32
So, γ is decreasing function on [0,4]. Therefore,
3 3x 2
G (x, 1) = − = ϕ (x) x ∈ [0, 4] .
4 256
Since
ϕ (0) ≤ 0
and
3
γ (4) = − <0
32
So, ϕ is decreasing function on [0, 4]. Therefore,
3
G (x, 1) = ϕ (x) ≤ ϕ (0) = < 1, x ∈ [0, 4] .
4
Case V (v): Let (x, y) ∈ (0, 4) × (0, 1). Then
∂G (4 − x)2 y
=− =0
∂y 32
if and only if
y = 0.
Hence,
∂G
(x, 0) = 0.
∂x
if and only if
x = 16 (1.33)
3 3x 2
ψ (x) = − x ∈ [0, 4] .
4 256
Since
ψ (0) ≤ 0
and
3
ψ (4) = − <0
32
So, ψ is decreasing function on [0,4]. Therefore,
9
ψ (x) ≥ ψ (4) = , x ∈ [0, 4] . (1.35)
16
Let e be a holomorphic function which is the solution of the differential equation
2ς e (ς )
= eς , ς ∈ D, e (0) = 0, e (0) = 1. (1.36)
e (ς ) + e (−ς )
This function e (ς ) satisfying (1.36) acts as the extremal function for the class
S∗s (e z ). This completes the proof of Theorem 1.4. This essentially completes the
proof of Theorem 1.6.
Definition 1.4 A function f (z), given by (1.1), is said to be in the class Cs (e z ) if
the following condition is satisfied:
2 z f (z)
≺ ez z ∈ U. (1.37)
( f (z) − f (−z))
Since
1
A2 Cs e z = .
4
By Lemma 1.2, we have the following theorem.
Theorem 1.7 If f ∈ Cs (e z ), then
15
≤ T2,1 ( f ) ≤ 1.
16
Both inequalities are sharp.
Now we find the bounds for T3,1 ( f ) . By adopting a similar technique as in
Theorem 1.6, we obtain the following theorem.
Theorem 1.8 If f ∈ Cs (e z ), then
253
≤ T3,1 ( f ) ≤ 1. (1.38)
288
1 The Second- and Third-Order Hermitian Toeplitz Determinants … 15
This function e (ς ) satisfying (1.39) acts as the extremal function for the class
Cs (e z ).
Acknowledgements The work of the third author is supported by a grant from the Science and
Engineering Research Board, Government of India, under Mathematical Research Impact Centric
Support of Department of Science and Technology (DST)(vide ref: MTR/2017/000607).
References
1. Ali, M.F., Thomas, D.K., Vasudevarao, A.: Toeplitz determinants whose elements are the
coefficients of analytic and univalent functions. Bull. Aust. Math. Soc. 97(2), 253–264 (2018)
2. Babalola, K.O.: On H3 (1) Hankel determinants for some classes of univalent functions. In:
Cho. Y.J. (ed.) Inequality Theory and Applications, vol. 6. pp. 1–7. Nova Science Publishers,
New York(2010)
3. Bansal, D., Maharana, S., Prajapat, J.K.: Third order Hankel determinant for certain univalent
functions. J. Korean Math. Soc. 52(6), 1139–1148 (2015)
4. Cho, N.E., et al.: Some coefficient inequalities related to the Hankel determinant for strongly
starlike functions of order alpha. J. Math. Inequal. 11(2), 429–439 (2017)
5. Cho, N.E., et al.: The bounds of some determinants for starlike functions of order alpha. Bull.
Malays. Math. Sci. Soc. 41(1), 523–535 (2018)
6. Cho, N.E., Kowalczyk, B., Lecko, A.: Sharp bounds of some coefficient functionals over the
class of functions convex in the direction of the imaginary axis. Bull. Aust. Math. Soc. 100(1),
86–96 (2019)
7. Cudna, K., et al.: The second and third-order Hermitian Toeplitz determinants for starlike and
convex functions of order α. Bol. Soc. Mat. Mex. (3) 26(2), 361–375 (2020)
8. Goodman, A.W.: Univalent functions. Mariner, Tampa (1983)
9. Janteng, A., Halim, S.A., Darus, M.: Hankel determinant for starlike and convex functions. Int.
J. Math. Anal. (Ruse) 1(13–16), 619–625 (2007)
10. Kowalczyk, B., Lecko, A., Sim, Y.J.: The sharp bound for the Hankel determinant of the third
kind for convex functions. Bull. Aust. Math. Soc. 97(3), 435–445 (2018)
11. Kowalczyk, B., et al.: The sharp bound of the third Hankel determinant for some classes of
analytic functions. Bull. Korean Math. Soc. 55(6), 1859–1868 (2018)
12. Kwon, O.S., Lecko, A., Sim, Y.J.: The bound of the Hankel determinant of the third kind for
starlike functions. Bull. Malays. Math. Sci. Soc. 42(2), 767–780 (2019)
13. Lecko, A., Sim, Y.J., Śmiarowska, B.: The sharp bound of the Hankel determinant of the third
kind for starlike functions of order 1/2. Complex Anal. Oper. Theory 13(5), 2231–2238 (2019)
14. Lee, S.K., Ravichandran, V., Supramaniam, S.: Bounds for the second Hankel determinant of
certain univalent functions. J. Inequal. Appl. 2013, 281, 17 (2013)
15. Mendiratta, R., Nagpal, S., Ravichandran, V.: On a subclass of strongly starlike functions
associated with exponential function. Bull. Malays. Math. Sci. Soc. 38(1), 365–386 (2015)
16. Pommerenke, C.: Univalent functions. Vandenhoeck and Ruprecht, Gottingen (1975)
17. Prajapat, J.K., et al.: Bounds on third Hankel determinant for close-to-convex functions. Acta
Univ. Sapientiae Math. 7(2), 210–219 (2015)
16 P. Gurusamy et al.
18. Robertson, M.I.S.: On the theory of univalent functions. Ann. Math. (2) 37(2), 374–408 (1936)
19. Sudharsan, T.V., Vijayalakshmi, S.P., Stephen, B.A.: Third Hankel determinant for a subclass
of analytic functions. Malaya J. Math. 2, 438–444 (2014)
Chapter 2
Some Results on a Starlike Class with
Respect to ( j, m)-Symmetric Functions
2.1 Introduction
We start with the following notations. The class of maps that are holomorphic on the
unit open disk = {ζ : ζ ∈ C with |ζ | < 1} and of form
∞
h(ζ ) = ζ + al ζ l (2.1)
l=2
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 17
P. V. Subrahmanyam et al. (eds.), Synergies in Analysis, Discrete Mathematics, Soft
Computing and Modelling, Forum for Interdisciplinary Mathematics,
https://doi.org/10.1007/978-981-19-7014-6_2
18 K. Renuka Devi et al.
Theorem 2.1 ([3]) For any mapping h : → C and an m-fold symmetric set,
there is a sequence of ( j, m)-symmetrical functions h j,m ,
m−1
h(ζ ) = h j,m (ζ ),
j=0
where
1 −ν j
m−1
h j,m (ζ ) = ε h(εν ζ ), (2.2)
m ν=0
(h ∈ A; m = 1, 2, ...; j = 0, 1, . . . , m − 1)
then
∞
h j,m (ζ ) = δl, j al ζ l , a1 = 1
l=1
1 (l− j)ν
m−1
1, l = lm + j;
δl, j = ε = (2.3)
m ν=0 0, l = lm + j,
Let S ∗(j,m) denote the subclass of S consisting of the mappings given by (2.1) such
that hζ j,m
h (ζ )
(ζ )
> 0, where h j,m (ζ ) is defined by (2.2) and ζ ∈ . Such functions
are said to be starlike w.r.to ( j, m)-symmetric functions.
∗( j,m)
Definition 2.3 A function h ∈ A of the form (2.1) is in class Ss (μ, η), if it
satisfies the following inequality:
2 Some Results on a Starlike Class with Respect … 19
ζ h (ζ )
− 1 < η μζ h (ζ ) + 1 ,
h (ζ ) h (ζ )
j,m j,m
1 − ζ λ(ζ )
G(ζ ) = , (2.5)
1 + μζ λ(ζ )
l−1
|lal − δl, j bl |2 ≤ 2(μη2 + 1) mδm, j |am ||bm |, (|a1 | = |b1 | = 1). (2.7)
m=1
Proof We shall follow the method given by Clunie et al. [1] and Thomas [6]. By the
Lemma 2.1, we have
ζ h (ζ ) 1 − ζ λ(ζ )
= ,
g j,m (ζ ) 1 + μζ λ(ζ )
1 − ζ λ(ζ )
ζ h (ζ ) = g j,m (ζ )
1 + μζ λ(ζ )
20 K. Renuka Devi et al.
or
[μζ h (ζ ) + g j,m (ζ )]ζ λ(ζ ) = g j,m (ζ ) − ζ h (ζ ).
Now, if
∞
ϕ(ζ ) = ζ λ(ζ ) = tl ζ l ,
l=1
then
|ϕ(ζ )| = η|ζ | f or ζ ∈ .
Therefore,
∞
∞ ∞
(μlal + δl, j bl )ζ l tl ζ l
= (δl, j bl − lal )ζ l . (2.8)
l=1 l=1 l=1
δl, j bl − lal = (μ + δ1, j )tl−1 + (2μa2 + δ2, j b2 )tl−2 + · · · + [(l − 1)μal−1 + δ(l−1), j bl−1 ]t1 .
Thus, the combination of coefficients on the r.h.s of (2.8) depends only on the
combination of coefficients (2μa2 + δ2, j b2 ), ..., [(l − 1)μal−1 + δ(l−1), j bl−1 ] on the
l.h.s. Further, for l ≥ 2,
l−1
l
(μ + δ1, j )ζ + (μmam + δm, j bm )ζ m ϕ(ζ ) = (δm, j bm − mam )ζ m
m=2 m=2
∞
+ cm ζ m (say). (2.9)
m=l+1
Squaring on both sides of Eq. (2.9), integrating over |ζ | = r < 1 and making use
of the relation |ϕ(ζ )| ≤ η|ζ |, we obtain
l ∞
|mam − δm, j bm |2 r 2m + |cm |2 r 2m
m=2 m=l+1
l−1
< η2 (μ + δ1,
2
j )r +
2
|μmam + δm, j bm |2 r 2m .
m=2
l
l−1
|mam − δm, j bm |2 < η2 (μ + δ1, j )2 + η2 |μmam + δm, j bm |2 ,
m=2 m=2
2 Some Results on a Starlike Class with Respect … 21
which implies
l−1
l−1
|lal − δl, j bl | ≤ η2 (μ + δ1, j )2 + η2 |μmam + δm, j bm |2 − |mam − δm, j bm |2 .
m=2 m=2
l−1
l−1
≤ η2 (μ + δ1, j )2 + (μ2 η2 − 1) m 2 |am |2 + (η2 − 1) δm, j |bm |2
m=2 m=2
l−1
l−1
+2μη2 |mam δm, j bm | + 2 |mam δm, j bm |,
m=2 m=2
or
l−1
l−1
|lal − δl, j bl | ≤ 2μη2 mδm, j |am ||bm | + 2 |mδm, j |am ||bm |,
m=1 m=1
l−1
|lal − bl |2 ≤ 2(μη2 + 1) m|am ||bm |, (|a1 | = |b1 | = 1).
m=1
Theorem 2.2 If h and g are in S and are such as in Lemma 2.2, then for l ≥ 2
1 1 1 1
|lal − δl, j bl |2 ≤ 2(μη2 + 1)C A(1 − , h j,m ) 2 A(1 − , g j,m ) 2 ,
l l
l−1
|lal − δl, j bl |2 ≤ 2(μη2 + 1) mδm, j |am ||bm |, (|a1 | = |b1 | = 1).
m=1
l−1 2 l−1 2
|lal − δl, j bl | ≤ 2μη
2 2
mδm, j |am | 2
mδm, j |bm | 2
m=1 m=1
l−1 2 l−1 2
+2 mδm, j |am | 2
mδm, j |bm | 2
m=1 m=1
l−1 2 l−1 2
2μη2
≤ 2l mδm, j |am |2 r 2m mδm, j |bm |2 r 2m
r m=1 m=1
l−1 2 l−1 2
2
+ 2l mδm, j |am | r 2 2m
mδm, j |bm | r2 2m
.
r m=1 m=1
2μη2 1 1 2 1 1
≤ A(r, h j,m ) 2 A(r, g j,m ) 2 + A(r, h j,m ) 2 A(r, g j,m ) 2 .
πr 2l πr 2l
∞
Since A(r, h j,m ) = π m=l−1 mδm, j |am |2 r 2m .
By choose r = 1 − 1
l
for l ≥ 2, we obtained the result.
Remark
∗( j,k)
Theorem 2.3 If h is in Ss (μ, η) with μη < 1, then al = O( 1l ) as l −→ ∞.
∗( j,m) ζ h (ζ )
Proof One can observe that, when μη < 1, for h ∈ Ss (μ, η), h j,m (ζ )
is bounded.
Now, we shall prove that
l−1
(l − δl, j )2 |al |2 < 2(1 − μη2 )(δ1, j + mδm, j |am |2 ) (|a1 | = |b1 | = 1).
m=2
∗( j,m)
If h ∈ Ss (μ, η) is given by (2.1), then by the Lemma 2.1
ζ h (ζ ) 1 − ζ λ(ζ )
= ,
h j,m (ζ ) 1 + μζ λ(ζ )
Now, if
2 Some Results on a Starlike Class with Respect … 23
∞
ζ ϕ(ζ ) = ζ λ(ζ ) = tl ζ l ,
l=1
then
|ϕ(ζ )| ≤ η|ζ | for ζ ∈ .
Therefore,
∞
∞
∞
∞
μζ + μlal ζ l − δl, j al ζ l ζ λ(ζ ) = δl, j al ζ l − ζ − lal ζ l .
l=2 l=1 l=1 l=2
This implies
∞
∞ ∞
(μl − δl, j )al ζ l tl ζ l
= (δl, j − l)al ζ l . (2.10)
l=1 l=1 l=1
We equate the coefficients of ζ l on both sides in the above equation and we obtain
(δl, j − l)al = (μ + δ1, j tl−1 + (2μ + δ2, j )a2 tl−2 + · · · + ((l − 1)μ + δ(l−1), j al−1 )t1 .
Thus, the combination of coefficients on the r.h.s of (2.10) depends only upon the
combination of coefficients
(2μ + δ2, j )a2 tl−2 , ..., ((l − 1)μ + δ(l−1), j al−1 )t1 .
∞
l
(μ − δ1, j ) + (μm − δm, j )am ζ m ϕ(ζ ) = (δm, j − m)am ζ m
m=2 m=2
∞
+ |cm |ζ m , (say). (2.11)
m=l+1
Upon squaring both sides of Eq. (2.11) and integrating over |ζ | = r < 1 and using
the relation |ϕ(ζ )| ≤ η|ζ |, we get
l ∞
|m − δm, j |2 |am |2 r 2m + |cm |2 r 2m
m=2 m=l+1
l−1
< η2 (μ − δ1, j )2 + |μm − δm, j |2 |am |2 r 2m .
m=2
24 K. Renuka Devi et al.
l
l−1
|m − δm, j | |am | ≤ η
2 2 2
(μ − δ1, j ) +2
|μm − δm, j |2 |am |2 .
m=2 m=2
This implies
⎡ ⎤
l−1
l−1
l − δl, j 2 |am |2 ≤ η2 ⎣(μ − δ1, j )2 + |μm − δm, j | |am | ⎦ −
2 2 |m − δm, j |2 |am |2
m=2 m=2
l−1
l−1
≤ η2 (μ − δ1, j )2 + (μ2 η2 − 1) k 2 |am |2 − 2(μη2 − 1) mδm, j |am |2
m=2 m=2
l−1
+ (η2 − 1) 2 |a |2 ,
δ1, j m
m=2
or
l−1
l − δl, j 2 |al |2 ≤ η2 (μ − δ1, j )2 − 2(μη2 − 1) mδm, j |am |2
m=2
l−1
≤ 2(1 − μη2 ) mδm, j |am |2 (|a1 | = |b1 | = 1) (2.12)
m=2
since μη < 1.
Now, we show that an = 0( 1l ) as l → ∞. From (2.12), we have
2
l−1
l − δl, j |al |2 < 2(1 − μη2 ) δ1, j + mδm, j |am |2 . (2.13)
m=2
Since ζ hhj,m(ζ(ζ) ) is bounded, h j,m (ζ ) is also bounded. Now, by following the method
given by Clunie [1], it follows that the area of the image of h j,k (ζ ) is
l−1
π δ1, j + mδm, j |am | 2
, (2.14)
m=2
and consequently, ∞
m=2 mδm, j |am | < ∞ and hence rl =
2 ∞
m=2 kδm, j |am |2 −→ 0
as l −→ ∞. Thus, we have
l−1
l−1
mδm, j |am |2 = (rm − rm+1 ) = r2 − rl = O(1) as l −→ ∞ (2.15)
m=2 m=2
2 Some Results on a Starlike Class with Respect … 25
Theorem 2.4 Let h(ζ ) ∈ A and in the form (2.1). If, for 0 ≤ μ ≤ 1, and 1
2
< η ≤ 1,
∞
(l + δl, j ) η(μl + δl, j )
+ |al | ≤ 1, (2.16)
l=2
η(μ + δ1, j ) − (1 − δ1, j ) η(μ + δ1, j ) − (1 − δ1, j )
∗( j,m)
then h(ζ ) ∈ Ss (μ, η).
∞
Proof Suppose that h(ζ ) = ζ + l=2 al ζ l . Then for |ζ | < 1,
∞
∞
ζ h (ζ ) − h j,m (ζ ) − ημζ h (ζ ) + h j,m (ζ ) = ζ + lal ζ l − δ1, j ζ − δ1, j al ζ l
l=2 l=2
∞
∞
−ημζ + μ lal ζ l + δ1, j ζ + δ1, j al ζ l
l=2 l=2
∞
∞
= (1 − δ1, j )ζ + (l + δl, j )al ζ l − η(μ + δ1, j )ζ + (μl + δl, j )al ζ l
l=2 l=2
∞
∞
≤ (1 − δ1, j )r + lδl, j )|al |r l − η (μ + δ1, j )r + (μl + δl, j )|al |r l
l=2 l=2
∞
∞
< (l + δl, j )|al | + (1 − δ1, j ) − η(μ + δ1, j ) + η(μl + δl, j )|al | r
l=2 l=2
∞
< (l + δl, j ) + η(μl + δl, j ) |al | + (1 − δ1, j ) − η(μ + δ1, j ) r
l=2
∞
< (l + δl, j ) + η(μl + δl, j ) |al | − η(μ + δ1, j ) − (1 − δ1, j )
l=2
∗( j,m)
so that h(ζ ) ∈ Ss (μ, η). Note that
for ζ = 1, 0 ≤ μ ≤ 1, 1
2
< η ≤ 1, l = 1, 2, 3, ...
The case of μ = 1 and η = 1 in Theorem 2.4, yields the following corollary, that
may provide ideas to several many problems involving coefficient techniques:
Corollary 2.2 With the hypothesis of Theorem 2.4, with μ = 1 and η = 1, we get
∞
(l + δl, j )|al | ≤ δ1, j .
l=2
Acknowledgements The work of Dr.S. Sivasubramanian is supported by a grant from the Sci-
ence and Engineering Research Board, Government of India, under Mathematical Research Impact
Centric Support of Department of Science and Technology (DST)vide ref: MTR/2017/000607.
References
1. Clunie, J., Keogh, F.R.: On starlike and schlit functions. J. London. Math. Soc. 35, 229–233
(1960)
2. Lakshminarasimhan, T.V.: On subclasses of functions starlike in the unit disc. J. Indian. Math.
Soc. 41, 233–243 (1977)
3. Liczberski, P., Polubinski, J.: On ( j, k)-symmtrical functions. Mathematica Bohemica 120, 13–
28 (1995)
4. Rabha M.D., Thomas, D.K.: Some subclasses of close-to-convex functions. J. Ramanujan Math.
Soc. 2(1), 85–100 (1987)
5. Sudharsan, T.V., Balasubrahmanyam, P., Subramanian, K.G.: On functions starlike with respect
to symmetric and conjugate points. Taiwanese J. Math. 2(1), 57–68 (1998)
6. Thomas, D.K.: On starlike and close-to-convex univalent functions. J. London. Math. Soc. 42,
472–435 (1967)
Chapter 3
Experimental Evaluation of Four
Intermediate Filters to Improve the
Motion Field Estimation
3.1 Introduction
Optical Flow estimation is one of the most challenging problems in computer vision.
Optical flow is defined as the per-pixel motion between two consecutive digital
images. Optical flow has many applications, such as video post-production, particle
velocimetry, video compression, control of autonomous vehicles, and many others.
In Fig. 3.1, we show an example of these applications. A transparent plastic model
full of water containing black plastic tracers is used to estimate the fluid’s velocity.
In Fig. 3.1, we show velocity estimation inside a process showing an oxygen inlet
and small black plastic tracers. Oxygen is injected from the left side of the plastic
model, causing the fluid to move. Tracer particles move at the same velocity as the
fluid. This application aims to determine the location where maximum velocity is
reached to predict erosion of internal walls.
In Fig. 3.2, we show the second application of the optical flow. We show two
consecutive images. In these images, we show a person riding a bike that moves to
the left. We estimated the optical flow between the two images; with this optical flow,
V. Lazcano (B)
Núcleo de Matemática, Física y Estadística, Facultad de Estudios Interdisciplinarios, Universidad
Mayor, Manuel Montt 318, Providencia, Santiago, Chile
e-mail: vanel.lazcano@umayor.cl
C. Isa-Mohor
Isa y Sabaj Ltda., Santiago, Chile
e-mail: ciisa@isaysabaj.cl
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 27
P. V. Subrahmanyam et al. (eds.), Synergies in Analysis, Discrete Mathematics, Soft
Computing and Modelling, Forum for Interdisciplinary Mathematics,
https://doi.org/10.1007/978-981-19-7014-6_3
28 V. Lazcano and C. Isa-Mohor
Fig. 3.1 Velocity fluid estimation inside a plastic model of a flow dynamic process. Optical flow
is represented with blue arrows in these two consecutive images
Fig. 3.2 Creation of an image in between. a Current image. b Next image. c Image in between
created based on optical estimation
we know the displacement for each pixel, so an image in between can be created.
We show in Fig. 3.2c an interpolation of two images using optical flow.
Since the seminal work of [1], many contributions have been made in order to improve
the optical flow estimation. In that work, the authors proposed a variational model
to estimate the optical flow. The proposal is an energy model to estimate the optical
flow estimation error, and the argument that minimizes that energy is the optical flow
of the sequence of images. The proposal is a model that uses a quadratic error, and
it means that the model is susceptible to outliers and the presence of noise. Zach et
al. [2] proposed another model based on the absolute value of the error. All those
models in an iterative way minimize the energy error model. In each iteration, some
of them filter the optical estimation to eliminate noise or outliers, avoiding noise and
outliers propagating across the iterations.
3 Experimental Evaluation of Four Intermediate Filters … 29
In [4], the authors integrated a Median filter in the optical flow model. Their proposal
formalizes the Median filter with a new model that integrates OF estimation over
a local neighborhood. In [5], the Median filter of size 3 × 3 is used to eliminate
irregularities of the optical flow and also noisy estimation. In [6], to enhance the OF
estimation, a bilateral filter is used. In that work, the authors replaced the anisotropic
diffusion of their proposed model with a novel multi-cue-driven bilateral filter that
considers the estimated occlusion.
Dérian et al. [7] utilizes another approach, where an optical flow model based
on wavelet analysis is presented. The multi-resolution approach used in the optical
flow estimation is similar to the multi-resolution used in wavelets analysis. The
authors constructed a scale-space representation of the optical flow; furthermore,
they provide a mechanism to locally approach the optical flow using high-order
polynomials by trunking wavelets at fine scales. This methodology was not evaluated
in a contemporaneous dataset but evaluated video sequences of moving fluids.
Motivation
In [8] an optical flow study is presented. This study consider the optical flow estima-
tion using image pyramid and also considers a theoretical analysis of warping, but
this study does not consider the study of intermediate filters. The study in [9] varies
the number of warpings, the image pyramid levels, and the influence of parameters
but does not take into account the intermediate filters. Those facts motivate us to
compare the optical flow estimation performance considering different intermediate
filters.
In this study, we use the optical flow estimation model proposed in [5]. The used
model is based on the absolute value of the optical flow estimation error (T V − L 1 )
30 V. Lazcano and C. Isa-Mohor
In Sect. 3.2, we explain the principal strategies used to estimate the optical flow. We
will briefly explain the linearization of the optical flow constraint, warping an image,
and image pyramid. We explain these strategies in order to make our manuscript
self-contained. In Sect. 3.3, we explain the filters considered in this work. In section
3.4, we explain experiments and used dataset. In Sect. 3.5, we present our obtained
results and a brief discussion about other methodologies and our results. Finally, in
Sect. 3.6 we present our conclusions and future work.
3.2 Preliminary
In order to state a model of the optical flow estimation, we consider two consecutive
color images I0 (reference image) and I1 (target image), where I0 , I1 : Ω → R3
and Ω a rectangular image domain; let u : Ω → R2 be the optical flow between
these two consecutive images (reference and target) where u(x) = (u 1 (x), u 2 (x))
has two components, that is to say, u 1 , u 2 : Ω → R. Optical flow estimation aims to
determine a motion field u(x) such that I0 (x) and I1 (x + u(x)) are measures of the
same pixel x, i.e.:
I0 (x) − I1 (x + u(x)) = 0. (3.1)
Equation 3.1 in the literature is called the color constancy constraint. Equation 3.1
is highly non-linear and a linearization is applied around a known optical flow u0 (x).
Considering a known value u0 (x) of the optical flow, a Taylor expansion is valid:
with , being the scalar product, I1 (x + u0 (x)) the warped image (I1 (x)) by a known
optical flow u0 (x), and ∇ I1 (x + u0 (x)) a gradient vector of I1 (x + u0 (x)).
3 Experimental Evaluation of Four Intermediate Filters … 31
(a ) (b )
( c) (d )
Fig. 3.3 Warping performed in a picture extracted from [11]. a Original image I (x). b Graphic
representation of the optical flow using orange arrows u0 (x). c Optical flow superimposed over the
original image. d Warped image by the optical flow or compensated image I (x + u0 (x))
Image warping is a process of image manipulation such that it distorts any shape
contained in the image. Given an image I (x) and an optical flow u0 (x), we can warp
the image I (x) to obtain I (x + u0 (x)) as we show in Fig. 3.3.
In large displacements, the Taylor approximation does not hold, and we use a coarse-
to-fine strategy. This strategy constructs a multi-scale pyramid down-sampling
images with a factor of 2. In Fig. 3.4, a scheme is shown of the image pyramid,
considering four levels.
Figure 3.4 begins with the coarsest scale, and each of them doubles the image
dimension in the previous one. At each level, the optical flow is filtered with an
intermediate filter to eliminate noise and outliers.
32 V. Lazcano and C. Isa-Mohor
Fig. 3.4 Image pyramid using an image of the sequence bandage. At the top, we have the coarsest
image and the finest at the bottom level. The computed optical flow in each level is used as an initial
condition for the optical flow estimation in the consecutive finer level
1
u i f (x) = w(x, y)u i (y), (3.3)
y∈N (x) w(x, y) y∈N (x)
3 Experimental Evaluation of Four Intermediate Filters … 33
with N (x) a neighborhood around x, w(x, y) are the exponential weights, and u i
(with i = 1, 2) are the vertical and horizontal components of the optical flow:
where φ S is the spatial distance between x and y, and φ I is the photo-metric distance
between the pixels in I0 (x) and I0 (y). φ S and φ I are Gaussian kernels given by
2 I0 (x)2
− x2 −
φ S (x) = e 2σS and, φ I (I0 (x)) = e 2σ I2
where σ S > 0, σ I > 0, and I0 (x) is the
reference image.
The Median filter is a non-linear filter, which is used to remove noise and outliers
from the optical flow estimation. Given a neighborhood around x (let’s say N (x)),
We sorted both the weights w and also the values in u(x). The auxiliary variables
are sw and su , representing weight values and optical flow values, respectively. We
found in the array sw positions p ∗ that hold
p∗
1
N
sw (i) < sw (i). (3.7)
i
2 i
34 V. Lazcano and C. Isa-Mohor
Finally, using the position p ∗ , we assign the weighted median filtered values of u:
u w f (x) = su ( p ∗ ). (3.8)
Our proposal is a weighted combination of the bilateral and the median. This lin-
ear combination has an adaptive weight α(x). This adaptive weight balances the
contribution of the bilateral and the median filter in the intermediate filtering.
Let u bl (x) and u m (x) be the filtered estimated optical flow by the bilateral filter
and the median filter, respectively. Following the ideas in [12], we constructed a
balance weight:
1
α(x) = , (3.9)
1 + e bl (x)−Dm (x))
(D
These two terms, Dbl (x) and Dm (x), represent the error of the OF in the point x
when the bilateral filter or the median filter are applied, respectively. The convex
combination gives the combined optical flow:
Depending on the values of αi (x), the flow can be more confident in the median filter
or the bilateral filter. Equation 3.9 shows that when Dbl >> Dm , αi value is almost
0 and when Dbl << Dm , α is almost 1.
This section presents the performed experiments and the dataset used to evaluate
optical flow estimation performance.
3 Experimental Evaluation of Four Intermediate Filters … 35
(a ) ( b)
(c) (d)
(e) (f )
(g ) ( h) ( i)
Fig. 3.5 Examples of the MPI-Sintel video dataset. a frame_0001 and b frame_0002 of video
sequence ambush_2. c Arrow representation of the ground truth OF of the video. d OF color-coded
representation. e frame_0001 and f frame_0002 images of the video sequence cave_4. g available
ground truth optical flow of two consecutive images. h Color-coded ground truth optical flow. i
Used optical flow color code
3.4.1 Dataset
The dataset contains different video sequences that present blur, fog, different illumi-
nation, and many scenes with large displacement and fast displacements. The dataset
is divided into two subsets. One set is a training set, and the second one is a validation
set. The training set is also divided into two subsets, one called clean and the other
called final. The final stage considers the effects mentioned above. This set is more
challenging than the clean one; therefore, we performed the experiments in this set.
Figure 3.5 shows examples of video sequences in the MPI-Sintel dataset, and we
offer the available ground truth optical flow with arrows and also using color code.
We show the used color code in Fig. 3.5i.
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Nerve enough
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.
Language: English
NERVE ENOUGH
By Richard Howells Watkins
“I’ll make that first jump; I’ve got to!” he cried, springing to his feet
and thumping a quick fist on the parachute packs. “You can’t trust
this fellow, and if he bungles it, we’re gone!”
“I’ll not bungle it,” retorted Burt Minster stubbornly. “And as for
nerve, I’ve more nerve than he has language, which is some.”
Jim Tyler slumped wearily against the side wall of the shack and
waited for the argument to subside.
“I stand ready to prove you a liar in any way you want to pick,”
Del O’Connell declared heatedly.
Burt Minster did not answer at once. His face reddened at the
challenge, but his eyes, as they dwelt upon the parachutes, were
merely thoughtful. Jim Tyler plunged into the lull.
“Since none of us has ever gone over, perhaps we’d better
rehearse a jump this evening, before we try it on the crowd,” he
suggested, in the hope that action would halt dissension.
But Burt Minster had by no means given up the controversy. He
had merely been planning.
“This Jenkins who is running the fair intimated to-day that he
might raise the ante if we pulled something particularly spectacular
the first day,” he said slowly. “And we need the money, if we’re ever
to get back where we started. Well, I have a scheme that’ll settle this
nerve question once and for all, and give us a big lift toward buying
another plane as well.”
“Out with it, then,” snapped Del O’Connell. “I’m willin’ already.”
Burt Minster laid a hand on the parachute packs.
“We have two of them, and we planned that the jumper should
wear both, as is customary. Well, instead of that, we’ll both jump, you
and I, at the same time.”
“And what would that prove?” snorted Del.
“I’m not through yet,” Burt rebuked him. “We’ll announce the thing
as a race to earth, the man landing first winning. You see, you don’t
have to pull the rip-cord that opens the parachute the minute you
leave the ship. You can fall free—an army expert fell almost two
thousand feet before he opened his ’chute—”
Del O’Connell’s eyes glinted.
“’Tis not a bad idea at all,” he admitted, and looked upon Burt
Minster with less rancor. “I like it fine.”
“Wait a minute,” interposed Jim Tyler. “You mean you’ll both jump,
and let yourselves fall a quarter of a mile or more? Why, that’s the
craziest—”
“And the man who pulls his rip-cord last wins, for he’ll land first,”
Del O’Connell explained. “As good a test of nerve as ever I heard
of.”
“Well, you can fly yourselves, then, for I’ll not have a hand in it,”
Jim Tyler announced firmly. “It isn’t necessary for you two to kill
yourselves to prove you’re fools. I’ll believe it now.”
His statement made no impression on his partners. This was no
sudden quarrel. Each, feeling guilty, was consequently touchy, and
doggedly set on doing his utmost to retrieve their misfortunes. And
from this attitude it was only a short step, in the ragged state of their
nerves, to an open conflict over the issue of courage—or any other
issue about which they could contend.
“Well, Jim,” said Burt Minster at last, as Tyler continued to stand
his ground unswervingly, “there’s another plane here at the fair, you
know. That fellow will take us both up if you won’t.”
Jim Tyler gave in at that, for he saw that his opposition to the plan
was only making them more eager to try it. Secretly he nursed the
hope that next day would bring them back to rational behavior.
But the opening hour of the fair found them still fixed in their resolve
to carry on perhaps the strangest duel of nerve that had ever been
devised. The three partners kept apart, since talk only led to
acrimony, and each at his post of observation watched the crowds
gathering.
They came in battered tin automobiles, and they came on foot,
and they came in ancient horse-drawn vehicles, from Baychester
County and from the county across the Baychester River which
flowed past the Fair Grounds. Jim Tyler’s airworn but still airworthy
Burgess training-plane was the center of a milling mob, for
Baychester was not so sophisticated as some of its neighbors, and a
flying machine was still an object of doubt and an object of awe. The
ropes about it strained under the pressure of the curious, and the
voices of the guards who reinforced the ropes grew hoarse and
querulous. And word of the race to the ground through the thin air
spread through the murmuring crowds.
The time of the flight came.
“Now boys, be sure and give us a good treat,” Jenkins, a stout,
harassed, badge-encrusted gentleman instructed, as he bustled up
to the shack wherein the partners had come together again.
“You’ll get it,” returned Burt Minster grimly.
“Two of them,” promised Del O’Connell, buckling the harness of
his ’chute about him, and taking a final glance at the dangling rip-
cord and the ring attached to it.
“I’ll make it worth your while,” the official declared, and dashed
away.
At the plane the three men waited, while space for a takeoff in the
infield was cleared of spectators. Jim Tyler warmed up his motor,
and then, throttling down, left the cockpit and confronted his
partners.
“If you’re set on going through with this fool thing I suppose I’ll
have to stand by,” he said briefly. “Where are you jumping from—
wing or cockpit?”
“Since we’re not pulling the rip-cords at once we might as well
jump from the cockpit,” said O’Connell. “You can signal to us better
from there and it will look more spectacular.”
“That suits me,” replied Burt Minster curtly.
“I won’t be able to get this bus up over six or seven thousand feet
with the weight of three men in her,” Jim calculated. “Suppose we
make it five thousand, to be sure?”
“A mile is plenty, since it’s going to be a sprint,” Del O’Connell
said, with a chuckle. “Though of course,” he added, looking sideways
at Minster, “one of us may not do much sprinting.”
“Speak for yourself,” growled the other man. “You’ll probably
starve to death before you get to the ground.”
“Remember, when I turn and put up five fingers, get ready,” Tyler
broke in hastily. “And when I nod, jump! One from each side. And
jump hard, so you’ll clear the tail.”
“Right,” assented Del O’Connell eagerly, and Burt Minster nodded
agreement.
The infield was clear at last. With a final glance at the fastenings
of their harness and the rip-cords that would release the parachutes,
the two men silently climbed into the rear cockpit. They wedged
themselves into the narrow seat. Then both turned automatically and
studied the direction and force of the wind, as revealed by the
whipping flags on the grandstand.
Jim Tyler gave the ship the throttle. Bouncing and lurching, it
charged into the wind, the propeller flickering as it cut the air and
flung it back upon the tense faces of pilot and ’chute jumpers. Far
across the infield the plane raced. Finally the wings took the burden
from the rubber-tired wheels. The ship, with a final jolt, parted
company with the ground, hung poised above the grass, and began
its upward climb.
Though it was an old story to them, the two men in the rear
cockpit looked downward, each upon his side, and the plane climbed
in great circles above the fair ground below. The green of the
countryside prevailed, but the brown of the oval racetrack cut
through it, and just outside this ellipse was a speckled band of many
indistinguishable colors that is the indication of people in masses.
Beyond that, behind the cigar-box grandstand, stretched a tightly
packed section of black and gray-black, where the automobiles of
the crowd were parked. Booths and buildings, gay with bunting,
displayed their tiny square outlines in regular patterns around the
ground.
And then, as the plane rose higher, the fair grounds contracted
until they were a mere detail of the landscape below—the great
green and brown squares and oblongs, with larger irregular patches
of woodland, interspersed here and there by tracts of well-watered
pasture land, of a lush green. Across it all, as if dividing all the world
into two parts, ran the almost straight course of the Baychester river.
Del O’Connell and Burt Minster at just the same time turned their
attention from the earth to the back of Jim Tyler’s head. They were
approaching their mark and both sensed it, although there was no
altimeter in their compartment.
The motor labored on, and both men thrust feet out straight, and
moved shoulders tentatively, as if to drive away any incipient
stiffness that might hinder action in that one swift leap into space.
Both were entirely at home in the air, as seamen are at home on the
water, but neither had ever gone out, deserting their craft for the
impalpable element in which it swam.
Suddenly Jim Tyler turned a grim face toward the rear cockpit and
raised his left hand, with fingers outstretched. Five thousand! For an
instant little Del O’Connell and big Burt Minster turned and looked at
each other. Determination was imprinted in the lines of both
countenances, and together they squirmed to their feet in that
cramped compartment, standing full in the buffeting stream of air
flung back by the whirling propeller. Del O’Connell, with an agile
twist, got one foot up on the rim of the cockpit and gripped the edge
with both his hands. His head turned forward, and his eyes fixed
themselves on the stern face of the pilot.
Burt, a little slower, slung a foot over his side of the machine, and
with one hand fumbled for the ripcord and dangling ring at the end of
it. Tyler nodded.
Del O’Connell, with a quick spring, brought his other foot up out of
the cockpit and, clinging with his hands, crouched on the edge of the
fuselage. His legs bent more sharply for the leap that would carry
him far out into space.
But just then the eyes of Jim Tyler caught a sudden flash of white
from the pack on Del’s back. The next instant the great silken
parachute whipped out of its confining envelop. Del’s rip-cord had
fouled on something inside the cockpit, and his eager jump to the rim
had jerked it.
The great spread of cloth billowed open instantly and whisked
backward in the grip of the wind. For just an instant Del, entirely
unconscious of what had occurred, held his place on the fuselage.
Then, like a stone from a catapult, he was whipped off his feet and
flung toward the tail of the racing plane.
The open parachute swept into the tail assembly. The
tremendous force of the wind ripped it from skirt to vent as it caught.
Shroud lines parted like threads. Then the silken cloth wrapped itself
about elevators, and several of the shrouds that did not snap
became entangled over the point of the balance of the rudder.
O’Connell’s whirling body struck the tail of the machine. Then it
swept past, dropping out into space. But the remaining shroud lines
were securely held by the rudder. O’Connell’s fall was checked by a
bone-jarring jerk. His body dangled below the tail of the plane,
swaying in the rush of the wind.
The plane wavered in the air, its flying speed dropping fast under
the resistance of the silken cloth whipping backward from the tail
assembly, and the drag of the man’s body swinging behind. Jim Tyler
opened the throttle full, and thrust the stick forward for a steep glide.
The elevators responded. They had been unhurt by the lashing
parachute. The nose of the plane turned earthward; its speed
increased.
The sudden catastrophe had come before Burt Minster had gone
over the side. He drew back in the cockpit and stared over at the
figure of Del O’Connell, dragging behind the plane by the precarious
strength of a few unsevered shroud lines. As he watched, he caught
sight of the white face of his partner, and saw that O’Connell, dazed
by the suddenness of the accident and his whip-like snap from the
cockpit, was just coming to a realization of what had occurred.
Jim Tyler turned and stared backward, too, and then the eyes of
Jim and Burt met. Speech was impossible in the fury of the motor’s
roar, but their eyes appealed to each other for help—for some way
out. The plane was diving sharply earthward; to check that dive
meant losing control of the ship; not to check it meant to crash at
terrific speed into the ground. There was no way of getting O’Connell
back into the ship; that was utterly impossible.
That communion of eyes lasted but a brief second; then both men
turned despairingly to the doomed man trailing behind the plunging
plane. They, too, were doomed in that headlong dash, but somehow
their plight seemed as nothing compared to his.
O’Connell had not lost his senses. They perceived that with both
hands he was fumbling, working at his right hip. Even as they
watched, his hand went to his left side in the same peculiar
movement. Then they comprehended.
O’Connell was unbuckling his harness. Already he had unclasped
the snap buckles that fastened the heavy webbing straps about his
thighs; now but one more buckle remained—the one across his
chest. He did not look toward the plane; his whole attention was
absorbed in his task, exceedingly difficult in that lashing wind,
dangling there in space at the end of the cords. But in an instant he
would no longer be dangling. The ship would be saved—at a price.
Jim Tyler watched, paralyzed by the horrible fascination of the
thing. In another instant O’Connell would have cast himself off from
the plane—and from life. His dry throat framed at last an inarticulate
sound of protest at the sight of that sacrifice. The wind swept it away
unheard.
Burt Minster, too, was watching. The breast buckle came apart.
Del O’Connell was free of the harness. He hung there by his hands,
and his face turned briefly toward them. A strained, twisted grin was
on it.
A pain shot through Jim Tyler’s shoulder; it was a blow from Burt
Minster’s heavy fist. The big man was squatting on top of the
fuselage.
“Right turn!”
His voice blared in the pilot’s ear, audible even above the thunder
of the motor. Jim obeyed automatically. The plane swerved sharply
to the right.
As the machine swung around, O’Connell’s body whipped
sidewise, no longer directly behind and below the tail. In that instant
Burt Minster leaped out into the air, all the strength of his powerful
muscles concentrated in the thrust of his legs. His body, its
momentum aided by the rush of air, shot through space. He crashed
like a plunging bull into the lean, small body of Del O’Connell.
The two men dropped together as the long arms of Burt wrapped
themselves about his partner.
The plane disappeared instantly from their view; they plunged
downward in a free drop, locked together, face to face. Air was all
about them; the thunder of the machine died away in their ears.
Beneath, the countryside was slowly expanding, opening up before
them like a magically blossoming flower.
“R-r-r-r-rip-cord!” roared Burt Minster. His own arms tightened
their clutch on Del O’Connell until the little man’s breath was
squeezed out of his chest. But even before Burt had spoken the
quick right hand of Del was wriggling downward, between Burt’s
shoulder and his own, toward the release ring. He found it. He
pulled.
Burt Minster’s breath followed Del O’Connell’s out of his body as
an iron band tightened across his breast; his thighs were squeezed
as if a boa had wrapped his constricting merciless folds about them.
Del felt a repetition of that shock that had hurled him from the
fuselage.
Burt emitted a sound, half expiration, half grunt. His parachute
had opened.
It spread above them like a shield. The country below ceased its
eerie expansion. Burt Minster’s grip about Del O’Connell’s chest
relaxed slightly, and the smaller man breathed again—deep, lung-
distending mouthfuls of sweet air. There was no longer any rush of
wind or roar of motor; nothing but a gentle, lulling sway from side to
side under that great canopy of silk.
Burt Minster spoke first.
“These things are supposed to handle up to four hundred pounds,
so I guess we’re all right,” he remarked, with an effort at a casual
tone.
Del blinked.
“If you’ll loosen up on those arms of yours, I’ll be able to get a grip
myself,” he answered. They adjusted their positions, and Del took
some of his weight from his hands by fastening his belt about Burt’s
harness. They continued to drift downward. The sudden cessation of
hubbub and speed made this gentle movement dreamlike.
Del O’Connell cleared his throat—and cleared it again. Finally he
muttered:
“That stuff about nerve, Burt—I’m a liar of the first water. Nerve?
You’re nothing else.”
“I saw what you were doing, yourself,” mumbled Burt Minster,
equally shamefaced and uncomfortable. “That certainly took guts,
Del.”
“I’m glad to be out of that mess,” said Del fervently. “Look! Here
comes Jim!”
Jim it was, and he was not above but below them. He was
climbing fast, and it was plain to see that he had complete control of
the ship. As they craned their necks toward the ascending plane he
banked sharply, and went circling under them, waving his hand
toward the tail. Nothing but a few tatters of silk and several shroud
lines trailed from the control surfaces of the tail assembly. Jim had
dived his encumbrance into ribbons.
With the plane whistling around them, they were wafted
downward almost directly over the fair grounds. A gentle wind was
drifting them toward it, for Jim had calculated well before signaling
for the jump. The earth was coming upward now with greater speed,
as their horizon drew in upon them. No longer could they survey half
the county.
Legs dangling, they waited. Past the eastern end of the racetrack
they drifted, and then, suddenly, the ground thudded up against their
feet, and down they went in a heap together. The parachute slipped
sideways, and lay billowing on the ground.
“We finished together, Del. It’s a dead heat,” said Burt Minster,
climbing to his feet and lifting the smaller man with him.
“Dead enough,” answered Del O’Connell emphatically. “But I’ve a
hunch this last little stunt has broken our run of bad luck, Burt. See!
Here comes Jenkins on the run, and I’m crashed if he hasn’t got his
checkbook in his hand!”
THE END
Updated editions will replace the previous one—the old editions will
be renamed.