Errata for Computer Vision: Algorithms and Applications

found by the Japanese translators of the Japanese edition

January 2, 2013
Hardcopy
p.#
12
22
32
32,33
34
34
35
36
48
53
63
92
96
96
105
111
112
117
121
124
132
134
138
140
156
158
160
169
172
188
203
212
277
281
282
285
290
291
291
291
307
309
309
310
313
314
315
317
317
319
321
333

Online
Sept. 3, 2010
13
24
35
35
37
37
38
39
54
59
70
105
109
109
119
126
127
133
138
142
150
157

181
192
195
212
232
240
313
314
318
320
323
329
330
330
330
348
350
350
351
354
355,356
358
359
359
362
364
378

Equation
Hardware impl.
Chap. 9
below (2.12)
above (2.12) and below (2.13)
Hier. of 2D trans.
above (2.22)
planar surface flow
above (2.23)
footnote 2
(2.79)
below (2.97)
3rd para. sec. 3.1.2
next para. below (3.9)
next para. below (3.9)
next para. below (3.28)
(3.39)
2nd para. before sec. 3.3.2
(3.50), Fig 3.24
(3.59)
next para. below (3.74)
2nd para. sec.3.5.3
3rd para. sec.3.5.3
para. in sec. 3.5.4 (below Fig.3.38)
last para. sec.3.5.4
below (3.102)
above (3.106)
(3.109)
1st para. sec.3.8
Ex.3.3
F
orsterHarris para.
(4.18)
(4.22)
footnote 3
(6.11) and above
below (6.26)
1st para. in Uncertainty modeling
below (6.37)
below (6.49)
(6.51)
below (6.51)
Figure 6.9
Figure 7.3
(7.18)
(7.20)
below (7.25)
footnote 8
above (7.35)
above (7.40)
below (7.45)
below (7.45)
3rd para. in 7.3.2
above (7.51), and (7.52), (7.55)
Ex 7.2

Correction
(see Section C.2) Appendix
example of computational photography
det(L) = 0 det(L) = 0 or |L| = 0
Hartley and Zisserman 2004, Chapter 2 Chapter 3
a subset of the more complex group subgroup
x
x
0 = H

x0 = Hx
0
y = a3 + a4 x + a5 y + a7 x2 + a6 xy a6 xy + a7 y 2
Hartley and Zisserman (2004, Section 2.4) Section 3.4
non-integral non-integer
x0 and y 0 in the right hand sides x
and y
e.g., from 1/125 to 1/250 from 1/250 to 1/125
Figure 2.32gi Figure 2.32fh
Figure 3.7d shows Figure 3.7e shows
in Figure 3.7e in Figure 3.7f
u
u
Gu
u
Gu
or u
u
G
remove =
remove ) of and 3.7.2).
ejx+ ej(x+)
LoG LoG
P (f ) Ps (f )
(Figure 3.33) (Figure 3.32)
the base-level Gaussian (L2 in Figure 3.34b) L1
The same applies to filtering the even sequence odd
Figure 3.40b illustrates ... 3.40d
where x = [...] is x = (...)
the posterior distribution for a given set of measurements y, p(y|x), p(x|y)
N N (i, j)
the IEEE International Conference on Image Processsing
remove ) from ... Sectoin 3.1.3) or ...
While Anandan (1989) and Kanade ...
kDA DB |
A DB k
kD
kDA DC |
kDA DC k
2
[ G ](x) I(x)]. [2 G ](x) I(x).
a different variance or certainty uncertainty
1
1
i
i
w(r) = (r)/r w(r) = (r)/r
(6.10) and 6.11) (6.10) and (6.11)
triangle triangle 4
Hartley and Zisserman 2004, Section 7.5 Section 8.5
...(xi cx )(xj cy ) + ... ...(xi cx )(xj cx ) + ...
Hartley and Zisserman 2004, Section 7.6 Section 8.6
(
xi , x
j ) (
xi , x
j )
x and 0 in the left camera image should be x0
add a period at the end of Eq.(7.18)
det |E 0 + (1 )E 1 | = 0 remove det
(Hartley and Zisserman 2004, p. 240) p. 259
Hartley and Zisserman (2004, p. 237) p. 256
Hartley and Zisserman (2004, p. 456) p. 472
weak perspective scaled orthographic
the 3 1 rows the 1 3 rows
... equal to sj and ... s2j
Cremers 2009). Another
ij ij
least media squares (LMS), least median of squares (LMS),

Hardcopy
p.#

Online
Sept. 3, 2010
389
390
391
405
411
419
430
434
442
455

Equation

630
646
691
691

(8.17)
(8.25)
(8.31)
(8.68)
below (8.71)
above sec. 8.5.2
2nd para. sec. 9.1
(9.10)
footnote 9
between (9.38) and (9.39)
(9.40)
Fig 10.5
(10.8)
below (10.18)
(10.35)
Fig 10.43
below (10.40)
Fig 10.48
5th para. sec. 10.5.2
Figure 11.3b
below (11.3)
Fig. 12.1
Fig. 12.8
2nd para. sec. 12.3.1
Fig. 12.14
Fig. 13.4
Fig. 13.8
footnote 12
below (14.34)
above (14.35)

624

711

sec.14.4.4 3rd para. from last

630
638
652
659
666
691

718
726
751,752
761
791

sec.14.5.2 3rd para.

Ex 14.5
(A.28)
(A.52)(A.54)
below (B.18)
Abdel-Hakim, ... (2006).

693

794

Arbel
aez, P., ... (2010).

699
703
707

797
802
808
813

Bartels, R. H., ... (1987)

Blake, ... (2010).
Capel, D. P. (2001).
Cootes, T. F. and Taylor, C. J. (2001).

710

817

Davis, T. A. (2008).

712
717

820
827

Diebel, J. (2006).
Foley, van Dam, Feiner, Hughes (1995).

719

829

Furukawa, Y. and Ponce, J. (2011).

721
725

832
837

725

838

729
736
736
748
742
753
762

843
853
853
870
861
876
889
899
931

Golovinskiy, A. ... (2006).

Hartley and Zisserman (2004)
Hastie, T., Tibshirani, R.,
and Friedman, J. (2001).
Huynh, D. Q., ... (2003).
Kumar, M. P., ... (2010)
Kumar, M. P., ... (2009)

344
356
361
368
378
381
400
400
416
430
449
450
451
454
459
471
474
508
516
518
524
548
553
566
607
607

792
references
references

475
482
489
512
514
514
517
524
538
541
590
592

Correction
eju eju
e2ju e2ju
I0 (s, ) I0 (s, )
wi,j wij
One popular choice for robust metrics in the L1 norm is
One example of ... 2004) which is the one One example of ... 2004) is the one
a planar surfaced surface
(h00 , h01 , h02 , h00 , h11 , h12 , h20 , h21 ) h10
1
1
ij ij
alpha (translucency) transparency
l = ... l(x) = ...
add missing (a)(b)(c)(d) from left to right.
... w(zi,j )[g(zi,j ) ...] w(zij )[g(zij ) ...]
The attenuation function I(x, y) (x, y)
F F
k k
/MI /M I
add missing (a)(b)(c)(d)(e) from left to right.
(Figure 10.55ab) ... (Figure 10.55cd) Figure 10.55 (or add (ab) in Fig.10.55)
x and 0 in the left camera image should be x0
(Hartley and Zisserman 2004, A5.2) A7.2
(ef) corresponding shape from shading ... (eh)
merged with their (weights) bottom left weights (bottom left)
(Section 3.7.1 (3.943.98) ) .5
add (a) (b) (c) (d) below each figure
add (c) below the right figure
... coordinates u and u0 u0 and u00
remove ( at the begining.
inverted file inverted index
inverted file index inverted index
using an active shape model (Cootes, Edwards, and Taylor 2001; ...
(Cootes, Cooper, Taylor et al. 1995
probabilistic latest semantic analysis (PLSA) latent
eye, nose, and face regions. mouth
|ai x bi |2 |aTi x bi |2
x
x

w(r) = (r)/r w(r) = (r)/r

color invariant characterstics color invariant characteristics
Submitted to PAMI. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-33(5):898916 (2011).
and Geeometric Modeling. Geometric
(2011). Markov Random Fields for Vision and Image Processing, MIT Press.
Super-resolution and Image Mosaicing Image Mosaicing and Super-resolution
In Medical Imaging. In SPIE Vol. 4322, Medical Imaging 2001: Image Processing, pp. 236248.
(2011). Algorithm 915, SuiteSparseQR: Multifrontal multithreaded rankrevealing sparse QR factorization. ACM Trans. on Mathematical Software, 38(1): Article 8.
Representing Attitude: Euler Angles, Unit Quaternions, and Rotation Vectors.
(1996) Computer Graphics: Principles and Practice. Addison-Wesley, Reading, MA, 2nd edition in C.
(2010). ... IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-32(8): 13621376.
ster, H. P., Pfister, H.,
Multiple View Geometry. 2nd Edition.
(2009). ... Springer-Verlag, New York, 2nd edition.
Outlier correcton Outlier correction
(2011). ... Journal of Machine Learning Research, 12(1):3167.
[Link], P. Torr, P. H. S.
Micus`k, ... (2008). Micusk
Two are the same reference and hence merged into (2006)
remove because this is an erroneous duplication of Kanade, T. and Okutomi, M. (1994).
Least median of squares regresssion regression
Subband Coding Subband Image Coding
Zitovaa, B. and Flusser, J. (2003). Zitov
a
ABCD20xx ABCD20xx
ABCD-20xx or ABCD-9x ABCD20xx or ABCD9x

Lischinski, D., ... (2006a) and (2006b)

Okutomi, M. and Kanade, T. (1994).
Rousseeuw, P. J. (1984).
Simoncelli, E. P. ... (1990b)