Professional Documents
Culture Documents
Download textbook Computer Vision Graphics And Image Processing Icvgip 2016 Satellite Workshops Wcva Dar And Medimage Guwahati India December 19 2016 Revised Selected Papers 1St Edition Snehasis Mukherjee Et Al Eds ebook all chapter pdf
Download textbook Computer Vision Graphics And Image Processing Icvgip 2016 Satellite Workshops Wcva Dar And Medimage Guwahati India December 19 2016 Revised Selected Papers 1St Edition Snehasis Mukherjee Et Al Eds ebook all chapter pdf
https://textbookfull.com/product/computer-vision-applications-
third-workshop-wcva-2018-held-in-conjunction-with-
icvgip-2018-hyderabad-india-december-18-2018-revised-selected-
papers-chetan-arora/
https://textbookfull.com/product/internetworked-world-15th-
workshop-on-e-business-web-2016-dublin-ireland-
december-10-2016-revised-selected-papers-1st-edition-ming-fan-et-
al-eds/
https://textbookfull.com/product/computer-vision-pattern-
recognition-image-processing-and-graphics-renu-rameshan/
https://textbookfull.com/product/autonomous-agents-and-
multiagent-systems-aamas-2016-workshops-best-papers-singapore-
singapore-may-9-10-2016-revised-selected-papers-1st-edition-
nardine-osman/
https://textbookfull.com/product/introduction-to-visual-
computing-core-concepts-in-computer-vision-graphics-and-image-
processing-1st-edition-aditi-majumder/
https://textbookfull.com/product/business-information-systems-
workshops-bis-2016-international-workshops-leipzig-germany-
july-6-8-2016-revised-papers-1st-edition-witold-abramowicz/
https://textbookfull.com/product/graph-drawing-and-network-
visualization-24th-international-symposium-gd-2016-athens-greece-
september-19-21-2016-revised-selected-papers-1st-edition-yifan-
Snehasis Mukherjee · Suvadip Mukherjee
Dipti Prasad Mukherjee · Jayanthi Sivaswamy
Suyash Awate · Srirangaraj Setlur
Anoop M. Namboodiri · Santanu Chaudhury (Eds.)
Computer Vision,
LNCS 10481
Graphics, and
Image Processing
ICVGIP 2016 Satellite Workshops, WCVA, DAR, and MedImage
Guwahati, India, December 19, 2016
Revised Selected Papers
123
Lecture Notes in Computer Science 10481
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7412
Snehasis Mukherjee Suvadip Mukherjee
•
Computer Vision,
Graphics, and
Image Processing
ICVGIP 2016 Satellite Workshops, WCVA, DAR, and MedImage
Guwahati, India, December 19, 2016
Revised Selected Papers
123
Editors
Snehasis Mukherjee Suyash Awate
Indian Institute of Information Technology Indian Institute of Technology Bombay
SriCity Mumbai
India India
Suvadip Mukherjee Srirangaraj Setlur
GE Global Research CEDAR
Bangalore Buffalo, NY
India USA
Dipti Prasad Mukherjee Anoop M. Namboodiri
Indian Statistical Institute International Institute of Information
Kolkata Technology
India Hyderabad
India
Jayanthi Sivaswamy
International Institute of Information Santanu Chaudhury
Technology CSIR-CEERI
Hyderabad Pilani, Rajasthan
India India
LNCS Sublibrary: SL6 – Image Processing, Computer Vision, Pattern Recognition, and Graphics
The tenth version of the Indian Conference on Vision, Graphics and Image Processing
(ICVGIP) was held at IIT Guwahati in December 2016. ICVGIP already taken a unique
position within the scientific community in India. Starting with the ninth version of
ICVGIP, workshops on contemporary topics have become an integral part of the
conference. The most recent in this series of ICVGIP workshops were held on
December 19, 2016, at the IIT Guwahati. Four tracks on Computer Vision Applica-
tions, Medical Image Processing, Document Analysis and Recognition, and Digital
Heritage were part of the workshop program this year. The collection of papers in this
edited volume is extended versions of papers presented in the first three of the four
above mentioned tracks. The authors of the papers presented in this workshop had
submitted an extended version of the paper keeping in mind the feedback received at
the workshop.
Overall 52 papers were submitted in the Computer Vision Applications track and
after a rigorous double-blind review process, we could accept 18 papers keeping the
quality and time-frame of a single-day workshop. This track mainly focused on
industrial applications of computer vision and related technologies.
The Medimage Workshop focused on problems in medical image computing and
invited original contributions to address these problems. The workshop received 14
papers and after a double-blind review process, 7 were accepted for presentation as
poster or oral.
The DAR Workshop was organized to bring together people working in the field of
Document Analysis and Recognition. The workshop received 10 submissions of which
7 papers were accepted for presentation, through a double-blind review process.
We would like to thank the Technical Program Committee members and the
reviewers for their time and dedication. We acknowledge support and guidance from
the dynamic Organizing Committee of ICVGIP 2016. In particular we appreciate
support received from Dr. Prithwijit Guha and Mathew Francis.
We hope that our efforts and the contributions presented in this collection will
encourage new ideas and further collaborative efforts between computer vision
researchers in industry and academia.
Editors
Snehasis Mukherjee IIIT SriCity, India
Suvadip Mukherjee GE Global Research, India
Dipti Prasad Mukherjee Indian Statistical Institute, India
Jayanthi Sivaswamy IIIT Hyderabad, India
Suyash Awate IIT Bombay, India
Srirangaraj Setlur CEDAR, SUNY Buffalo, USA
Anoop M. Namboodiri IIIT Hyderabad, India
Santanu Chaudhury CSIR-CEERI Pilani, India
WCVA
DAR
MedImage
1 Introduction
Document images are used as proof for authentication and business transactions.
Traditionally digital watermarking has been used as a primary technique for copyright
protection and integrity management of document images [1–3]. The document image
consists of information with various levels of sensitivity. For instance, in a cheque
image, the signature and amount are dynamically changing information for each che-
que and thus possess highest level of sensitivity. The bank name, logo, cheque number
contain regeneratable information content and hence constitute lower level of sensi-
tivity. There also exists many empty areas in a cheque which can be classified as
insensitive areas. Each sensitivity level needs different type of protection. Therefore,
there is a need to use multiple watermarking techniques on the different areas of the
same document image. The multiple watermarking schemes have two fold objectives:
improve the perceptual quality of the watermarked image by reducing embedding
capacity; perform tamper detection and recovery with better accuracy.
2 Literature Review
3 Proposed Model
The proposed model for the novel intelligent multiple watermarking system consists of
two processes namely Embedding and Extraction. The Embedding process divides the
input document image into blocks and intelligently determines the type of water-
marking to be applied for each block. The watermarking algorithm depends on the
information content of the image. This is primarily available through the energy
component and hence luminance component in transformation is used. Further image is
converted back to color after watermarking to produce watermarked image. The
embedding technique depends on the type of watermarking. Robust watermarking is
implemented using integer wavelet embedding [23] and fragile watermarking is
accomplished using contourlet based embedding [24]. Extraction process is carried on
the blocks of the watermarked image. The result of the watermark extraction depends
on the type of the watermarking. The outcome of the robust watermark extraction on
the block of the watermarked image is content authentication of the block. The out-
come of fragile watermark extraction on the block of the watermarked image results in
tamper detection and recovery of information content in the block. The following
subsection explores the embedding process and extraction process in detail.
document image. Experiments have been conducted on all document images corpus to
analyze the effect of size of the block on accuracy in identification of the type of the
block and processing time. For each block, gradient binarized version of the infor-
mation content in the block is obtained. Further, the sensitivity level of each block and
type of watermarking required is found automatically. Subsequently, appropriate
watermark embedding algorithm is applied for each block.
The gradient binarized version of the information content in the block is computed
using the following algorithm:
In this algorithm, the values of the weights w1 and w2 is empirically set to 0.5. The
number of iterations required for termination of this algorithm depends on the distri-
bution of the information content in the block. The outcome of this algorithm is a
binary version of the block that gives segmentation of foreground and background
information contents in the block.
A Novel Intelligent Multiple Watermarking Schemes 7
Experiments have been conducted on the document image corpus to decide on the
appropriate range to relative energy distribution and homogeneity values for deter-
mining the sensitivity levels of the blocks. The average REDb and HMb values for
different types of information content in these document images is calculated and
values are recorded in Table 1. It can be observed from the values in Table 1 that RED
values for blocks of document image containing dynamically changing information
content are in the range 0.7–0.85 and HM values lie between 0.29–0.50. Thus, sen-
sitivity level of the block with HM less than 0.5 and RED above 0.7 is set to 0.
Similarly, it can see in Table 1 that blocks of the document image containing pre-
printed information content has RED values above 0.3 and HM values in between 0.5
to 0.85. Therefore, sensitivity level of these blocks is set to 1. For all the other blocks,
sensitivity level is set to 2.
The type of watermarking used depends on the sensitivity levels of the information
content in the block. Highly sensitive blocks are protected using fragile watermarking
technique. In this paper an effective fragile watermarking technique based on con-
tourlets [24] is used. Partially sensitive blocks are protected using robust watermarking
technique [23]. The size of the block is decided based on two factors: effectiveness in
the identification of the sensitivity of the block and processing time for identification.
8 K.R. Chetan and S. Nirmala
Table 1. Computation of RED and HM values for different classes of document images in the
corpus
Document Preprinted Dynamically changing
image class information information
RED HM RED HM
Cheques 0.33 0.67 0.72 0.50
Bills 0.36 0.82 0.79 0.35
Identity cards 0.62 0.85 0.76 0.39
Marks cards 0.30 0.79 0.85 0.29
Certificates 0.54 0.51 0.70 0.36
Experiments have been conducted exhaustively on all the document images to measure
the impact of size of the block against accuracy in identifying sensitivity level of the
block. The average number of blocks expected for each sensitivity level and number of
blocks being accurately identified is recorded in Table 2. It can be observed from
average accuracy in identification values that the blocks of lesser size exhibits higher
accuracy.
Table 2. Impact of size of the blocks of a document image on accuracy in identification of its
sensitivity level and processing time for identification
Block size Type-0 Type-1 Type-2 IA (in %) Processing time (in secs)
EB IB EB IB EB IB
32 32 348 339 210 204 466 466 98.53 102.12
64 364 83 75 57 52 116 116 94.92 78.7
128 3128 22 20 17 15 25 25 93.75 32.11
256 3256 5 4 4 3 7 7 87.50 21.16
and generated is compared and if there is a mismatch, the corresponding block of the
document image is declared “inauthentic”. However, if there is a match, then water-
mark extraction is carried out based on the type of the block. The extracted and
generated watermarks are compared for similarity using Feature Similarity Index [24]
and based on the comparison, the tamper detection of the block is decided. If the block
is tampered, recovery of information content is made by extracting watermark
embedded at robust locations [24]. During robust watermark extraction, the watermark
is extracted from the LL-band of the integer wavelet transformation performed on the
block of a document image. The extracted watermark is decoded using binary block
coding technique [23]. The decoded watermark is compared with original watermark
and decision of content authentication of the block is performed [23].
4 Results
We have created a corpus of document images. All the images in the corpus are
scanned document images. The classes of document image corpus considered are
Cheques, Bills, Identity Cards, Marks cards and Certificates. Each class of document
image consists of 30 images. The results of the identification of type of the blocks of a
sample document image in the corpus are shown in Fig. 2.
Fig. 2. Results of identification of the type of the block of a sample Cheque image
It can be observed in Fig. 2, that there are three types of blocks in the sample
Cheque image. The blocks with dashed border are Type-0 blocks. They are highly
sensitive blocks containing large variations in the information content and distribution
of the information. The blocks with dotted border are Type-1 blocks i.e. partially
sensitive blocks which contain preprinted information. They have moderate homo-
geneity in distribution of the information content. Remaining type of blocks in the
document image are the insensitive blocks (Type-2) which contain less energy and
10 K.R. Chetan and S. Nirmala
Fig. 3. Results of proposed multiple watermarking system (a) source document image
(b) watermarked image (c) original watermark for robust watermarking (d) zoomed up Partially
sensitive block with salt and pepper noise attack (e) zoomed up Highly sensitive tampered block
(f) extracted robust watermark (g) tamper detection results (h) tamper recovery result
5 Analysis
using Peak Signal to Noise Ratio (PSNR) [25]. A graph of PSNR values is depicted in
Fig. 4 for different classes of the document images. The graph shown in Fig. 4 reveals
that PSNR values of the multiple watermarking schemes are better than robust and
fragile watermarking schemes when applied separately. This increase in PSNR and
subsequently the perceptual quality of the watermarked image is due to the fact that all
the blocks of the document image are not watermarked. The quantity of the watermark
to be embedded depends on the type of the block. Hence, the noise induced due to
watermarking is reduced to some extent and this result in the better fidelity of the
watermarked image.
Table 4. Average TDA and TRA values for different intentional attacks
Intentional attacks Existing Fragile water Proposed multiple
marking scheme [24] watermarking scheme
TDA TRA TDA TRA
Insertion 0.9 0.87 0.91 0.90
Deletion 0.92 0.91 0.92 0.92
Modification 0.87 0.87 0.90 0.89
6 Conclusions
A novel intelligent multiple watermarking schemes are proposed in this paper. The
blocks of a document image have been automatically classified into various sensitivity
levels with greater accuracy. The performance analysis of the proposed approach
reveals improvement in the perceptual quality of the watermarked image. The proposed
scheme also outperforms the existing methods [23, 24] in providing robustness, tamper
detection and recovery capabilities. Improvement on the accuracy of identification of
type of block is taken up as future work of the current study.
References
1. Wu, M., Liu, B.: Watermarking for image authentication. In: Proceedings of the IEEE
International Conference on Image Processing, pp. 437–441 (1998)
2. Cox, I., Miller, M., Bloom, J., Fridrich, J., Kalker, T.: Digital Watermarking And
Steganography. Morgan Kaufmann Publishers Inc., San Francisco (2007)
3. Hartung, F., Kutter, M.: Mutimedia Watermarking Techniques. Proc. IEEE 87(7), 1079–
1107 (2002)
A Novel Intelligent Multiple Watermarking Schemes 13
4. Potdar, V.M., Han, S., Chang, E.: A survey of digital image watermarking techniques. In:
3rd IEEE International Conference on Industrial Informatics, pp. 709–716 (2005). doi:10.
1109/Indin.2005.1560462
5. Mirza, H., Thai, H., Nakao, Z.: Color image watermarking and self-recovery based on
independent component analysis. In: Rutkowski, L., Tadeusiewicz, R., Zadeh, L.A., Zurada,
J.M. (eds.) ICAISC 2008. LNCS, vol. 5097, pp. 839–849. Springer, Heidelberg (2008).
doi:10.1007/978-3-540-69731-2_80
6. Wang, M.S., Chen, W.C.: A majority-voting based watermarking scheme for color image
tamper detection and recovery. Comput. Stand. Interfaces 29, 561–571 (2007)
7. Bas, P., Chassery, J.M., Macq, B.: Geometrically invariant watermarking using feature
points. IEEE Trans. Image Process. 11(9), 1014–1028 (2002)
8. Qi, W., Li, X., Yang, B., Cheng, D.: Document watermarking scheme for information
tracking. J. Commun. 29(10), 183–190 (2008)
9. Dawei, Z., Guanrong, C., Wenbo, L.: A chaos-based robust wavelet-domain watermarking
algorithm. Chaos, Solitons Fractals 22(1), 47–54 (2004)
10. Schirripa, G., Simonetti, C., Cozzella, L.: Fragile digital watermarking by synthetic
holograms. In: Proceedings of the European Symposium on Optics/Fotonics in Security &
Defence, London, pp. 173–182 (2004)
11. Houmansadr, A., et al.: Robust content-based video watermarking exploiting motion entropy
masking effect. In: Proceedings of the International Conference on Signal Processing and
Multimedia Applications, pp. 252–259 (2006)
12. Kankanhalli, M.S., Ramakrishnan, K.R.: Adaptive visible watermarking of images. In: IEEE
International Conference on Multimedia Computing and Systems, vol. 1, pp. 568–573
(1999)
13. Radharani, S., et al.: A study on watermarking schemes for image authentication. Int.
J. Comput. Appl. (0975 – 8887) 2(4), 24–32 (2010)
14. Kay, S., Izquierdo, E.: Robust content based image watermarking. In: Proceedings of the
Workshop on Image Analysis for Multimedia Interactive Services (2001)
15. Kim, M.-A., Lee, W.-H.: A content-based fragile watermarking scheme for image
authentication. In: Chi, C.-H., Lam, K.-Y. (eds.) AWCC 2004. LNCS, vol. 3309,
pp. 258–265. Springer, Heidelberg (2004). doi:10.1007/978-3-540-30483-8_31
16. Habib, M., Sarhan, S., Rajab, L.: A Robust-Fragile dual watermarking system in the DCT
domain. In: Khosla, R., Howlett, R.J., Jain, L.C. (eds.) KES 2005. LNCS, vol. 3682,
pp. 548–553. Springer, Heidelberg (2005). doi:10.1007/11552451_74
17. Shieh, C.-S., et al.: Genetic watermarking based on transform-domain techniques. J. Pattern
Recogn. 37, 555–565 (2004)
18. Goldberg, D.E.: Genetic Algorithms in Search Optimization and Machine Learning.
Addison-Wesley, Reading (1992)
19. Lu, Z.-M., Xu, D.-G., Sun, S.-H.: Multipurpose image watermarking algorithm based on
multistage vector quantization. IEEE Trans. Image Process. 14(6), 822–831 (2005). doi:10.
1109/Tip.2005.847324
20. Sheppard, N.P., Safavi-Naini, R., Ogunbona, P.: On multiple watermarking. In: Dittmann, J.,
Nahrstedt, K., Wohlmacher, D. (eds.) Multimedia and Security: New Challenges Workshop,
p. 38871 (2001)
21. Voloshynovskiy, S., Pereira, S., Pun, T., Eggers, J.J., Su, J.K.: Attacks on digital
watermarks: classification, estimation based attacks, and benchmarks. IEEE Commun. Mag.
39(8), 118–126 (2001)
22. Wang, S., Zhang, X.: Watermarking scheme capable of resisting sensitivity attack. IEEE
Signal Process. Lett. 14(2), 125–128 (2007)
14 K.R. Chetan and S. Nirmala
23. Chetan, K.R., Nirmala, S.: An efficient and secure robust watermarking scheme for
document images using integer wavelets and block coding of binary watermarks. J. Inf.
Secur. Appl. 24–25, 13–24 (2015)
24. Chetan, K.R., Nirmala, S.: A novel fragile watermarking scheme based on contourlets for
effective tamper detection, localization and recovery of handwritten document images. IEEE
Signal Process. Lett. (Communicated)
25. Aggarwal, E.D.: An efficient watermarking algorithm to improve payload and robustness
without affecting image perceptual quality. J. Comput. 2(4) (2010). ISSN 2151-9617
26. Zhu, X., et al.: Normalized correlation-based quantization modulation for robust water-
marking. IEEE Trans. Multimed. 16(7), 1888–1904 (2014)
Experimental Evaluation of 3D Kinect Face
Database
1 Introduction
Facial expressions, poses and variations have attracted the research world since long, as
it is easily obtainable and convenient biometric trait as compared to the iris, voice, gait
etc. 2D facial images has well defined roots in the world of biometric research due to
low cost of its acquisition system and wide availability [1]. But 2D face recognition
system faces its limitations when it comes to mostly illumination and pose variation [2].
In order to overcome these short comings of 2D recognition, 3D recognition system
captured the market as security concern has increased from local to the defense level.
Research in 3D biometric was an expensive task as the expense of system requirement
for acquiring 3D images was very high and time consuming [3] until the development
of efficient, low-cost RGB-D Kinect camera. This system provides 2D RGB image as
well as depth information i.e. distance from each pixel to the sensor [4].
Images captured by Kinect camera has low resolution and noisy yet it has more
spatial information compared to 2D in form of depth which is a robust inherent
property associated for 3D face recognition against uncontrolled environment.
Ekenel et al. obtained 3-D face recognition approach using the discrete cosine trans-
form (DCT) which is a local appearance-based model at feature level [5]. Huynh et al.
has proposed a new LBP based descriptor namely Gradient-LBP (G-LBP) for gender
recognition task on EURECOM and Texas database [6]. Vezzetti et al. proposed a new
3D face recognition algorithm, whose framework based on extracting facial landmarks
using the geometrical properties of facial shape [7]. Min et al. have generated a 3D
database based on the Kinect sensor having 52 subjects over two sessions for 2D, 2.5D,
3D and video. Here recognition rates are calculated for 2D, 2.5D, and 3D-based face
data using standard face recognition techniques like PCA, LBP, SIFT, LGBP, ICP, and
TPS and also RGB and Depth images were fused using score-level fusion [8]. Ajmera
et al. has computed CRR based on modified SURF descriptors and image enhancement
techniques and filters like adaptive histogram equalization, NLM filter etc. for their
internal database and has compared it with Eurecom and Curtin face database and also
has performed scored level fusion [9]. Hg et al. had proposed RGB-D Face dataset
(VAP database) of 31 subjects containing 1581 images and have developed a face
detection protocol using curvature analysis technique and tested for VAP database [10].
Goswami et al. had generated IIIT-D RGB-D face database of 106 subjects with
multiple Images per subject. Authors has also proposed an algorithm for 3D face
recognition which involves computation of entropy map and visual saliency map
followed by HOG descriptor for feature extraction and finally the use of Random
Decision Forest (RDF) classifier for establishing identity. The algorithm was tested for
IIIT-D and Eurecom Database [4, 11]. Table 1 is giving the brief idea about existing
Kinect base 3d facial databases.
Mao et al. has work on identification and filling of expansion holes. Here the holes
are identified based on depth histogram and the filling of holes is done using linear
interpolation and graph based interpolation method [13]. Solh and AlRegib has pro-
posed two approaches for dis-occlusion removal in Depth Image-Based Rendering
(DIBR): hierarchical hole-filling (HHF) and depth adaptive hierarchical hole-filling.
The said approaches follows pyramid like approach from lower resolution estimate of
3D wrapped image to estimate the said hole pixels [14]. Wang et al. have propose a
hole filling algorithm to improve image quality of DIBR. Here to determine the order of
hole filling the depth information is added to the priority calculation function. Next
when searching for the best matching block the gradient information is used as aux-
iliary information [15, 16]. Feng et al. has proposed an adaptive background biased
depth map hole-filling method [17]. Based on this literature survey we are proposing
simple weighted average nonlinear interpolation hole/patch removal algorithm for the
3D database generated at our laboratory. We are presenting our GU-RGBD facial
database having variation in pose, expressions and occlusion; collected over two ses-
sions for 64 subjects. This database will be made available in public domain for
research purpose. Extensive experimental evaluation is performed for the State of the
Art face recognition algorithm in the form of identification and verification rate.
The rest of the paper is organized as follows: Sect. 2 describes the GU-RGBD data
generation setup and protocol. Section 3 is giving the detail explanation of the pro-
posed filter. Section 4 explains the experimental evaluation protocol for database.
Section 5 comprises of results and discussion and the final conclusion is given in
Sect. 6.
2 3D Database Generation
For every subject, eight RGB and eight depth images were captured to form total of
16 in every session. The session 1 consists 16 images captured in controlled envi-
ronmental condition and session 2 consist of 16 images captured in uncontrolled
environmental condition by keeping window open for ambient light to enter the lab and
the QTH sources were off. Thus total of 32 images captured per subject. The
GU-RGBD database has the enrollment of 64 subjects out of which 49 are males and
15 are females from various age group. The size of database is 64(subjects) * 32
(images per subject) = 2048 images. The sample images of database are shown in
Fig. 2.
The images captured by Kinect are noisy and inaccurate [8]. The presence of patches
(zero value pixels present on the depth face) degrade the captured information and thus
affects the recognition rate and hence it has to be enhanced in the pre-processing stage.
We are using the interpolation method to replace the patch with the neighboring pixels.
There are various scenarios of the characteristics of patch like localized position on
facial triangle, size of patch and patch position on the boundaries. Hence sometimes
surrounding information if not available in such scenarios especially when patch is
placed at boundary positions. Hence, there is a need to allocate the weightage for the
populace of surrounding pixels in the form of some constants i.e. a1, a2. The proposed
filter acts as the solution for patches present on depth faces.
The depth images of M N dimensions usually having the 4:3 aspect ratio are
extended by adding M/4 dummy rows and column pixels respectively for the higher
dimensions of aspect ratio as shown in Fig. 3(a). This is extended so as to avoid the
occurrence of computational errors for the pixels at the outer boundary as window of
filter expands. Then ‘NaN’ values are assigned to the dummy rows and column pixels
so as to avoid filling of false information for patches in vicinity of boundary. The
proposed filter scan image by initially setting a l l kernel i.e. ‘u(l,l)’ to locate zero
value pixel (which is a patch) to be replaced.
Fig. 3. Filter implementation: (a) schematic view of filter, (b)–(d) patches of different sizes
Another random document with
no related content on Scribd:
however, the more he washed in the hot tears, the redder grew his
clothes, until he was just the colour of the scarlet bean blossom.
“You told me a story,” said Gillydrop to the giant when he saw how
red he was getting.
“I know I did,” said Dunderhead, drying his eyes, for he had now
wept enough, and was growing hungry; “but if I hadn’t told you a
story, I wouldn’t have got any supper. You’ll never be green again, so
don’t trouble your head. I’m going to get some wood to cook these
nice fat children.”
On hearing this, Teddy and Tilly roared like bulls, and Gillydrop
roared too, for he was afraid he would never be able to go back to
Faeryland in his red clothes; but the giant only laughed at them, and
went out to light a fire under his big kettle.
Gillydrop was naturally very cross with the giant for having deceived
him, and determined to punish him for having done so. Bringing the
two children to Dunderhead for his supper could not be the kindly
deed he had to do, or else he would have turned green again; so
Gillydrop made up his mind to take Teddy and Tilly back to earth,
and thus leave Dunderhead without his supper. While he was thus
making up his mind, seated at one end of the table, the two children,
seated at the other end, were crying bitterly at the plight in which
they now found themselves, for it certainly is not a nice thing to be
boiled for an ogre’s supper.
“Poor mother!” wailed Tilly, weeping; “she’ll miss us so much.”
“I don’t know if she will,” replied Teddy dolefully; “we’ve always been
so naughty, I daresay she’ll be glad we’ve gone.”
“Oh no, she won’t,” said Tilly, nodding her head; “she loves us too
much for that; but if we could get back I’d be so good.”
“And so would I,” cried Teddy; and then they both wept again, while
Gillydrop, seeing their tears, wept also out of sheer sympathy.
“Perhaps the giant will only eat one of us,” said Tilly after a pause;
“so while one of us is boiling, the other must run away and go back
to comfort mother.”
“Who will be boiled?” asked Teddy sadly. “Will you, Tilly?”
“I don’t like being boiled,” answered Tilly, with a shudder. “I’m sure it
isn’t nice.”
“Well, I don’t like being boiled either,” observed Teddy. “Suppose we
draw lots who is to run away.”
“Yes, that would be fair,” said Tilly, drying her eyes; “and the one who
wins must go back to cheer mother.”
Gillydrop was quite sorry now that he had brought them for
Dunderhead’s supper, when he heard how they regretted their
mother; so he made up his mind to save them.
“You shall neither of you be boiled,” he said, walking up to them
across the table, which was like a large plain. “I will take you back to
your mother.”
“But how?” asked Teddy and Tilly, both together. “We cannot go back
across the sea alone.”
“Oh yes, you can,” replied the Red Elf. “I brought you here, and can
send you back; that is, if I only had a leaf.”
“Here is one,” cried Tilly eagerly, pulling a faded leaf out of her
pocket. “I picked it up in the wood to-day, it had such pretty red and
yellow colours.”
“Oh, that will do for a boat,” said Gillydrop joyfully.
“But it’s so small,” objected Teddy.
“I’ll make it large enough,” said the elf. “You’ll see.”
“But how can we go on without sails or oars?” said Tilly timidly.
“You don’t need any,” rejoined Gillydrop, laughing; “you know every
tree has power to draw back its own leaves. The boat we came in
was a leaf, and, as soon as it was launched on the air, it went
straight back to the tree in the Country of the Giants upon which it
had grown; and as this leaf comes from a tree on earth, it will go
straight back to its tree.”
“Then we can get home,” cried Tilly, clapping her hands, “for the tree
isn’t far from mother’s cottage.”
“Mind, you are never to be naughty again,” said Gillydrop solemnly.
“Oh, no, no!” cried both children.
“And be very, very good to your mother.”
“Yes, yes! We’ll be very good.”
“Then go down to the beach by the path,” said Gillydrop, spreading
his wings. “I’ll fly down and get the boat ready; be quick, or the giant
will return.”
Then he flew away through the open window, and Teddy scrambled
down the steep path, followed by Tilly, both of them in a great fright
lest the giant should catch sight of them and pop them into his big
kettle. When they reached the beach, they found Gillydrop had
launched the leaf, which had now been transformed into a beautiful
red and yellow coloured boat.
“Good-bye,” said Gillydrop, as soon as they were comfortably seated
in the boat. “I’m sorry I brought you here, but it will do you no harm,
as it will teach you to be good. Mind you don’t quarrel in the boat—if
you do, the leaf will vanish, and you’ll sink for ever in the black
waves.”
“Oh, we’ll be very, very good,” promised both the children eagerly,
and then Gillydrop gave the boat a push, so that it moved rapidly
away from the land, leaving him seated on the beach, a lonely little
red figure.
Teddy and Tilly were rather afraid at finding themselves alone in the
darkness, but they kissed one another, and fell asleep, while the
leaf-boat sailed rapidly over the Sea of Darkness towards its parent
tree. When the children awoke, they found themselves lying on the
ground under the tree, and there above them was their red and
yellow boat, hanging, a red and yellow leaf, on a high bough.
“Now we’ll go home,” cried Tilly, jumping up; “now we’ll go home to
mother.”
“And be very good,” said Teddy, also rising.
“Yes; very, very good,” replied Tilly. And then, taking one another’s
hands, they ran home to their cottage through the dark forest.
Dame Alice, who thought they had lost themselves in the wood, was
very glad to see them, and, after she had kissed them, gave them a
good supper of bread and milk, which they enjoyed very much, for
you see they were very hungry with the long journey.
They told Dame Alice all their adventures, and she was very glad
they had gone to the Giants’ Country, for she guessed, like the wise
mother she was, that this was the lesson the faeries had foretold.
Ever afterwards, Teddy and Tilly were good children; there never
were two such good children, because they thought, if they were not
good, they would be taken back to the Giants’ Country and boiled for
an ogre’s supper. But after a time they liked to do good actions
because they found it pleasant, and Dame Alice was so pleased with
their behaviour that she made a rhyme about them, which soon
passed into a proverb:
“The magic power of a faery
Cures a child when quite contrary.”
III.
When Gillydrop saw the magic boat disappear into the darkness of
the sea, he thought that, now he had done one kindly deed, his
clothes would change from red to green, and he would be able to
return to his dear Faeryland. But nothing of the sort occurred, and
the poor elf began to cry again, thinking he was lost for ever, but this
time his tears were not red, which was a good sign, although he did
not know it.
Very soon he heard Dunderhead roaring for the loss of his supper,
so, drying his eyes, he flew back again to the hall of the castle, to
see what the giant was doing. He found a great fire was lighted, over
which was suspended a great kettle filled with water, which was now
boiling hot. Dunderhead was searching everywhere for the children,
and when he saw Gillydrop he shook his great fist at him.
“Where’s my supper, you red rag?” he roared fiercely.
“Your supper has gone back to earth,” replied Gillydrop angrily, for no
one likes to be called a red rag. “You told me a story, so I thought I’d
punish you.”
“Oh, did you?” bellowed Dunderhead, in a rage. “Then I’ll punish you
also for spoiling my supper.” And before Gillydrop could fly away, he
caught him in his great hand and popped him into the boiling water.
Oh, it was terribly hot, and Gillydrop thought it was all over with him;
but, being a Faery, he could not be killed, as the foolish giant might
have known. He sank down, down, right to the bottom of the great
kettle, and then arose once more to the top. As soon as he found his
head above water, he sprang out of the kettle and flew away high
above the head of Dunderhead, who could only shake his fist at him.
To his delight and surprise, Gillydrop found his clothes had all
changed from red to green, and instead of being dressed in crimson,
his suit was now of a beautiful emerald colour. He was so delighted
that he flew down on to the floor of the hall, and began to dance and
sing, while the giant joined in as he tried to catch him; so that they
had quite a duet.
Gillydrop. Now I’m gay instead of sad,
For I’m good instead of bad:
Dreadful lessons I have had.
Giant. I will catch and beat you!
But you see he could not do that, because Gillydrop was too quick
for him, and flew round the hall, laughing at Dunderhead, who roared
with anger. Then the elf flew out on to the terrace which overlooked
the Sea of Darkness, followed by the giant. Gillydrop flew down on to
the beach to escape the ogre, and Dunderhead tried to follow; but,
as he could not fly, he fell right into the Sea of Darkness. Dear me!
what a terrible splash he made! The waves arose as high as the
castle walls, but then they settled down again over Dunderhead, who
was suffocated in the black billows. He was the very last of the
giants, and now his bones lie white and gleaming in the depths of the
Sea of Darkness, where nobody will ever find them—nor do I think
any one would trouble to look for them.
As for Gillydrop, now that Dunderhead was dead, he flew away
across the dreary plain towards Faeryland, and soon arrived at the
borders of the sullen grey sea which still rolled under the pale light of
the moon. Gillydrop was not a bit afraid now, because his clothes
were green once more, and he had performed one kindly deed; so
he sat down on the seashore and sang this song:
“When from Faeryland I fled,
All my nice clothes turned to red;
Now in emerald suit I stand—
Take me back to Faeryland.”
And as he sang the grey ocean faded away, and in its place he saw
the green trees of the faery forest, waving their branches in the silver
moonlight. Only a bright sparkling stream now flowed between
Gillydrop and Faeryland; so, spreading his silver and blue wings, he
flew across the water, singing gaily:
“Thanks, dear Oberon. At last
All my naughtiness is past;
Home I come without a stain,
And will never roam again.”
So at last Gillydrop got back to Faeryland after all his trials, and ever
afterwards was one of the most contented elves ever known. You
may be sure he never wanted to see the Country of the Giants
again, and whatever King Oberon said he did willingly, because he
knew it must be right.
He was quite a hero among the faeries, and had the honour of telling
all his adventures to King Oberon himself, which he did so nicely that
the King gave him a title, and ever afterwards he was called “Sir
Gillydrop the Fearless.”
SHADOWLAND
IT was Christmas Eve, and the snow, falling heavily over a great city,
was trying to hide with its beautiful white robe all the black, ugly
houses and the narrow, muddy streets. The gas lamps stood up
proudly, each on its tall post, and cast their yellow light on the
crowds of people hurrying along with their arms filled with many
lovely presents for good children.
“They are poor things,” said the gas lamps scornfully. “If we did not
shed our light upon them, they would be lost in the streets.”
“Ah, but the people you despise made you,” cried the church bells,
which were calling the people to prayer. “They made you—they
made you, and gave you your beautiful yellow crowns.”
But the street lamps said nothing, because they could not deny what
the church bells said, and instead of acknowledging that they owed
all their beauty to the people they despised, remained obstinately
silent.
Near one of these lamp-posts, at the end of a street, stood a
ragged boy, who shivered dreadfully in his old clothes, and stamped
about to keep himself warm. The boy’s name was Tom, and he was
a crossing-sweeper, as could be seen by his well-worn broom. He
was very cold and very hungry, for he had not earned a copper all
day, and the gaily-dressed army of people swept selfishly past him,
thinking only of their Christmas dinners and warm homes.
The snowflakes fell from the leaden-coloured sky like great white
angels, to tell the earth that Christ would be born again on that night,
but Tom did not have any such ideas, as he was quite ignorant of
angels, and even of the birth of the child-Christ. He only looked upon
the snow as a cold and cruel thing, which made him shiver with pain,
and was a great trouble to brush away from his crossing.
And overhead the mellow bells clashed out
their glad tidings in the bitterly chill air, while
below, in the warm, well-lighted churches, the
organ rolled out its hymns of praise, and the
worshippers said to one another, “Christ is
born again.”
But poor Tom!
Ah, how cold and hungry he was, standing in
the bright glare of the lamp, with his rags drawn closely round him for
protection against the falling snow. The throng of people grew
thinner and thinner, the gaily-decorated shops put up their shutters,
the lights died out in the painted windows of the churches, the bells
were silent, and only poor Tom remained in the deserted, lonely
streets, with the falling snowflakes changing him to a white statue.
He was thinking about going to his garret, when a gentleman,
wrapped in furs, passed along quickly, and just as he came near
Tom, dropped his purse, but, not perceiving his loss, walked on
rapidly through the driving snow. Tom’s first idea was to pick the
purse up and restore it to its owner, whom Tom knew very well by
sight, for he was a poet, who daily passed by Tom’s crossing. Then
Tom paused for a moment as he thought of all the beautiful things
the money in that purse would buy; while he hesitated, the poet
disappeared in the darkness of the night, so Tom was left alone with
the purse at his feet.
There it lay, a black object on the pure white
snow, and as Tom picked it up, he felt that it
was filled with money. Oh, how many things of
use to him could that money buy—bread and
meat and a cup of warm coffee—which would
do him good. Tom slipped it into his pocket, and thought he would
buy something to eat; but just at that moment he seemed to hear a
whisper in the air,—
AS TOM PICKED IT UP HE FELT THAT IT WAS
FILLED WITH MONEY
ONCE upon a time, long long years ago, there was a shepherd
called Duldy, who dwelt in the forests which clothed the base of the
great mountain of Kel. This mountain was in the centre of an
immense plain, watered by many rivers, and dotted over with many
cities, for the kingdom of Metella was a very rich place indeed, so
rich that the inhabitants looked upon gold in the same way as we
look upon tin or iron, as quite a common thing. The plain was very
fertile by reason of the great rivers which flowed through it like silver
threads, and all these rivers took their rise in the mountain of Kel, a
mighty snow-clad peak which shot up, white and shining, to the blue
sky from amidst the bright green of its encircling forests.
There were old stories handed down from father to son, which said
that the mountain was once a volcano, which, breathing nothing but
fire, sent great streams of red-hot lava down to the fertile plain, to
wither and blight all the beautiful gardens and rich corn-fields. But
the fires in the breast of the mountain had long since died out, and
for many centuries the black, rugged summit had been covered with
snow, while countless streams, caused by the melting of the glaciers,
fell down its rocky sides, and, flowing through the cool, green pine
forests, spread themselves over the thirsty plain, so that it bloomed
like a beautiful garden.
Duldy lived in these scented pine forests, and
was supposed to be the son of an old couple
called Dull and Day, from whence by joining
both names he got his own Duldy; but he was
really a lost child whom old Father Dull had
found, seventeen years before, on the banks
of the Foam, one of the bright sparkling
streams which flowed from the snowy heights
above. Dull took the child home to his wife
Day, who was overcome with joy, for she