This action might not be possible to undo. Are you sure you want to continue?

2, 2010

**Effective MSE optimization in fractal image compression
**

A.Muruganandham

Sona College of Technology, salem-05.,India. muruga_salem@rediffmail.com,

**Dr.R.S.D.wahida banu
**

Govt Engineering College Salem-11, India rsdwb@yahoo.com

Abstract- The Fractal image compression encodes image at low bitrate with acceptable image quality, but time taken for encoding is large. In this paper we proposed a fast fractal encoding using particle swarm optimization (PSO). Here optimization technique is used to optimize MSE between range block and domain block. PSO technique speedup the fractal encoder and preserve the image quality. Keywords- mean square error (MSE) ,particle swarm optimization (PSO), fractal image compression (FIC), Iteration Function System (IFS)

I. INTRODUCTION The idea of the fractal image compression (FIC) is based on the assumption that the image redundancies can be efficiently exploited by means of block self-affine transformations. The fractal transform for image compression was introduced in 1985 by Barnsley and Demko [1,2]. The first practical fractal image compression scheme was introduced in 1992 by Jacquin [3,4]. One of the main disadvantages using exhaustive search strategy is the very high encoding time. Therefore, decreasing the encoding time is an interesting research topic for FIC [3, 4]. An approach for decreasing encoding time is using the stochastic optimization methods such as genetic algorithm (GA). Some recent GA-based methods are proposed to improve the efficiency [5, 6]. The idea of special correlation of an image is used in these methods, which is of great interesting. While the chromosomes in GA consist of all range blocks which leads to high encoding speed and particular properties of natural images have never been used that will results in lose of visual effect in the retrieved image. Other researchers focused on improvements of the search process to make it faster by tree structure search methods [12,13], parallel search methods [14,15] or quad tree partitioning of range blocks [9,16,]. In this paper, we present a fast fractal encoding using particle swarm optimization. The outline of the remaining part of this paper is as follows: Section II includes fractal image coding. Section III involves implementation of PSO. Section IV concerns the proposed fast fractal encoder using PSO, and in Section V experimental results are included. In Section VI, we present our conclusions.

II. FRACTAL IMAGE COMPRESSION ALGORITHM The Iteration Function System (IFS) is the fundamental idea of fractal image compression in which the governing theorems are the Collage Theorem and the Contractive Mapping Fixed-Point Theorem [7]. The encoding unit of FIC for given gray level image of size N x N is (N/L)2 of nonoverlapping range blocks of size L x L which forms the range pool R. For each range block v in R, one search in the (N - 2L + 1)2 overlapping domain blocks of size 2L x 2L which forms the domain pool D to find the best match. The parameters describing this fractal affine transformation of domain block into range block form the fractal compression code of v. The parameters of fractal affine transformation Φ of domain block into range block is domain block coordinates-(tx, ty), Dihedral transformation-d, contrast scaling-p, brightness offset-q. Flowchart for this fractal affine transformation is illustrated in fig. 1 and it is given by

x y u ( x, y )

a11 a12 a 21 a 22 0 0

0 0 p

x y u ( x, y )

tx ty , q

(1)

a11

where the 2 x 2 sub-matrix Dihedral transformations in (2)

T0 1 0 0 1 , T1 1 0 0 1 ,

a12 a 22

is one of the

a 21

T2

1 0 , 0 1

T3

1 0

0

, T4 1

0 1 1 0

,

T5

0 1 , 1 0

T6

0 1

1 , 0

T7

0 1

1 0

(2)

238

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

Input image of size NxN Partition input image into (N/L) non overlapping range blocks of size L x L and (N-2L+1)2 overlapping domain blocks of size 2L x 2L

2

(IJCSIS) International Journal of Computer Science and Information Security, Vol. 8, No. 2, 2010

pk

L2 u k , v

L 1 i 0

L 1 j 0

u k (i, j )

L 1 i 0 L 1 j

L 1 i 0

L 1 j 0 2

v(i, j ) ,

L2 u k , u k

u (i, j ) 0 k

(5)

For Range blocks 1 to (N/L) 2

qk

For Domain blocks 1 to (N-2L+1)2

1 L2

L 1 i 0

L 1 j 0

v(i, j )

pk

L 1 i 0

L 1 j 0

u k (i, j ) .

(6)

For Orientation 1 to 8

4.

Calculate Scaling, Offset and Mean Square Error

As u runs over all of the domain blocks in D to find the best match, the terms tx and ty can be obtained together with d and the specific p and q corresponding this d, the affine transformation (1) is found for the given range block v.

Store quantized scaling and offset, orientation and position of the domain block of minimum MSE

In practice, tx, ty, d, p, and q can be encoded using log2(N), log2(N), 3, 5, and 7 bits, respectively, which are regarded as the compression code of v. Finally, the encoding process is completed as v runs over all of the (N/L)2 range blocks in R. Fig. 2 show the MSE vs. quantization parameter for an randomly selected range block of size 8 x 8 from 256 x 256 Lena image. From fig. 2, choosing 5 bits and 7 bits as quantization parameter for scale and offset value is justified.

16 scale bits = 3 4 5 6 7 8 9 10 11

Fig. 1 fractal affine transformation of domain block into range block The above parameters are found using the following procedure 1. 2. the domain block is first down-sampled to L x L and denoted by u The down-sampled block is transformed subject to the eight transformations Tk:k = 0,. . . ,7 in the Dihedral on the pixel positions and are denoted by uk, k = 0,1, . . . ,7, where u0 = u. The transformations T1 and T2 correspond to the flips of u along the horizontal and vertical lines, respectively. T3 is the flip along both the horizontal and vertical lines. T4, T5, T6, and T7 are the transformations of T0, T1, T2, and T3 performed by an additional flip along the main diagonal line, respectively. For each domain block, there are eight separate MSE computations required to find the index d such that

14

12

MSE

10

8

6

4

3

4

5

6 7 8 9 no. of bits for Offset quantization

10

11

3.

Fig. 2 MSE Vs. quantization parameter To decode, chooses any image as the initial one and makes up the (N/L)2 affine transformations from the compression codes to obtain a new image, and proceeds recursively. According to Partitioned Iteration Function Theorem (PIFS), the sequence of images will converge. According to user’s application the stopping criterion of the recursion is designed. The final image is the retrieved image of fractal coding. III. PARTICLE SWARM OPTIMIZATION (PSO) PSO is a population-based algorithm for searching global optimum. It ties to artificial life, like fish schooling or bird flocking, and has some common features of evolutionary

d

arg min{ MSE (( p k u k

q k ), v) : k

0,1,...,7}

(3)

where

MSE (u , v)

1 L2

L 1

(u (i, j

i, j 0

0) v(i, j )) 2

(4)

Here, pk and qk can be computed directly as

239

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

computation such as fitness evaluation. The original idea of PSO is to simulate a simplified social behavior [8, 9]. Similar to the crossover operation of the GA, in PSO the particles are adjusted toward the best individual experience (PBEST) and the best social experience (GBEST). However, PSO is unlike a GA in that each potential solution, particle is ‘‘flying” through hyperspace with a velocity. Moreover, the particles and the swarm have memory; in the population of the GA memory does not exist.

**(IJCSIS) International Journal of Computer Science and Information Security, Vol. 8, No. 2, 2010
**

(Start)

Initialization

Fitness evaluation Experience updating

Let xj,d(t) and vj,d(t) denote the dth dimensional value of the vector of position and velocity of jth particle in the swarm, respectively, at time t. The PSO model can be expressed as

v j ,d (t )

v j ,d (t 1) c1. 1.(x*,d j c2 . 2 .(x

# d

x j ,d (t 1))

(7)

Moving

No

x j ,d (t 1)),

Stopping criterion reached? Yes

x j ,d (t )

where

x j ,d (t 1) v j ,d (t ),

* j ,d

(Stop)

(8)

Fig. 3 Block diagram of PSO IV. PROPOSED METHOD In the proposed fast fractal encoding using PSO, we reduce the encoding time by reducing the searching time to find a best match domain block for the given range block from all domain blocks. Flowchart of the fractal encoding of the proposed method is shown in fig. 4.

x

**(PBEST) denotes the best position of jth
**

# xd

particle up to time t-1 and (GBEST) denotes the best position of the whole swarm up to time t-1, φ1 and φ2 are random numbers, and c1 and c2 represent the individuality and sociality coefficients, respectively. The population size is first determined, and the velocity and position of each particle are initialized. Each particle moves according to (7) and (8), and the fitness is then calculated. Meanwhile, the best positions of each swarm and particles are recorded. Finally, as the stopping criterion is satisfied, the best position of the swarm is the final solution. The block diagram of PSO is displayed in Fig. 3 and the main steps are given as follows: 1. 2. Set the swarm size. Initialize the velocity and the position of each particle randomly. For each j, evaluate the fitness value of xj and update the individual best position found. 3.

Input image of size NxN

Partition input image into (N/L)2 non overlapping range blocks of size L x L and (N-2L+1)2 overlapping domain blocks of size 2L x 2L

x *,d j

For Range blocks 1 to (N/L) 2

**if better fitness is
**

Calculate scale, offset, orientation and position of the domain block of minimum MSE from all domain blocks using PSO

Find the new best position of the whole swarm. Update the swarm best position x# if the fitness of the new best position is better than that of the previous swarm. If the stopping criterion is satisfied, then stop. For each particle, update the position and the velocity according (8) and (7). Go to step 2.

4. 5.

Store quantized scaling and offset, orientation and position of the domain block of minimum MSE

Fig.4 Fractal encoding of proposed method Domain block of minimum MSE is found by PSO using the steps given below:

240

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

1.

Set the swarm size proportional to (N2L+1)2/(maximum no. of iterations for PSO) and initialize the velocity and the position of each particle randomly Fitness value includes finding MSE between domain block specified by the particles position and given range block using eqn. (3) Update swarm best position if the fitness of the new best position is better than that of the previous swarm. If swarm best position is not changed for some percentage of maximum iteration for PSO, then stop The best position of the particles is updated using eqn. (7) and (8) and goto step 2. V.

XPERIMENTAL RESULTS

(IJCSIS) International Journal of Computer Science and Information Security, Vol. 8, No. 2, 2010

2.

Fig. 7 shows the variation in PSNR by varying the stopping criterion in fast fractal encoding using PSO by changing the percentage of maximum iteration of PSO. From fig.7 variation of PSNR with variation of stopping criterion is very less. Hence a 10% of maximum iteration for PSO is choose as stopping criterion.

34.34 34.338 34.336 34.334 PSNR (dB) 34.332 34.33 34.328 34.326 34.324 10

3. 4. 5.

The fast fractal encoding using PSO results have been compared to the full search FIC mentioned in the previous sections in terms of encoding time and PSNR. Fig. 5 shows the original image Lena 256 x 256 at 8bpp. Fig. 6 shows the decoded Lena image using full search FIC and fast fractal encoding using PSO. The numeric results containing bitrate, encoding time and PSNR of decoded image for various images are tabulated in table I.

20 30 40 50 60 70 Stopping criterion (percentage of maximum iteration for PSO)

80

Fig. 7 variation in PSNR by varying the stopping criterion

Fig. 4. Original 256 x 256 Lena image

Fig. 4. Original 256 x 256 Lena image

(a) (b) Fig. 5. Decoded Lena image at 0.5 bits per pixel for (a) Full search FIC (b) proposed fast fractal encodind using PSO

(a)

(b)

Fig. 5. Decoded Lena image at 0.5 bits per pixel for (a) Full search FIC (b) proposed fast fractal encodind using PSO

241

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

(IJCSIS) International Journal of Computer Science and Information Security, Vol. 8, No. 2, 2010

VI. CONCLUSION Fractal image compression can produce better compression ratio at acceptable quality. By using PSO for fractal coding we can reduce the encoding time with 1.2dB loss in image quality. VII. REFERENCES

[1] M.F. BARNSLEY, S. DEMKO, ITERATED FUNCTION SYSTEMS AND THE GLOBAL CONSTRUCTION OF FRACTALS, PROC . ROY. SOC. LOND A399 (1985) 243–275. A.E. Jacquin, Fractal image coding: a review, Proc. IEEE 10 (1993)1451–1465. A.E. JACQUIN, IMAGE CODING BASED ON A FRACTAL THEORY OF ITERATED CONTRACTIVE IMAGE TRANSFORMATIONS, IEEE TRANSACTIONS ON SIGNAL PROCESSING 1 (1992) 18–30. M.F. Barnsley, A.D. Sloan, A better way to compress images, BYTE Magazine (1988) 215–233 M. POLVERE, M. N APPI, SPEED-UP IN FRACTAL IMAGE CODING: COMPARISON OF METHODS, IEEE T RANSACTIONS ON IMAGE PROCESSING 9 (2000) 1002–1009. T.K. TRUONG, J.H. JENG, I.S. REED, P.C. LEE, A.Q. LI, A FAST

ENCODING ALGORITHM FOR FRACTAL IMAGE COMPRESSION USING THE DCT INNER PRODUCT, IEEE T RANSACTIONS ON IMAGE PROCESSING 9 (4) (2000) 529–535.

[2] [3]

[4] [5] Fig. 4. Original 256 x 256 Lena image [6]

[7]

[8]

M.S. WU, J.H. JENG, J.G. HSIEH, SCHEMA GENETIC ALGORITHM FOR FRACTAL IMAGE COMPRESSION, E NGINEERING APPLICATIONS OF ARTIFICIAL I NTELLIGENCE 20 (2007) 531–538. M.S. WU, W.C. TENG, J.H. JENG, J.G. HSIEH, SPATIAL

CORRELATION GENETIC ALGORITHM FOR FRACTAL IMAGE COMPRESSION, CHAOS SOLITONS & FRACTALS 28 (2) (2006) 497–

[9] [10]

(a) (b)

[11] Fig. 5. Decoded Lena image at 0.5 bits per pixel for (a) Full search FIC (b) proposed fast fractal encodind using PSO TABLE I NUMERIC RESULTS [12] [13]

Input image

Bitrate (bpp)

Method Full search FIC Proposed method Full search FIC Proposed method Full search FIC Proposed method

Encoding Time (hh:mm:ss) 09:07:20 00:15:34 09:02:12 00:17:37 09:02:49 00:15:24

PSNR (dB) 35.80 35.03

[14]

Lena

0.5

[15]

[16]

33.64 32.77 35.11 34.23

510. Y. FISHER, FRACTAL IMAGE COMPRESSION—THEORY AND APPLICATION. NEW YORK: SPRINGER -VERLAG, 1994. J. KENNEDY, R.C. EBERHART, PARTICLE SWARM OPTIMIZATION, IN: PROCEEDINGS OF IEEE INTERNATIONAL C ONFERENCE ON NEURAL NETWORKS, PERTH, AUSTRALIA, VOL. 4, 1995, PP. 1942– 1948. R.C. EBERHART, J. KENNEDY, A NEW OPTIMIZER USING PARTICLE SWARM THEORY, IN: PROCEEDINGS OF IEEE INTERNATIONAL SYMPOSIUM ON MICRO MACHINE AND HUMAN SCIENCE, NAGOYA, JAPAN, 1995, PP. 39–43. B. BANI-EQBAL, SPEEDING UP FRACTAL IMAGE COMPRESSION, PROC . SPIE: STILL-IMAGE COMPRESS. 2418 (1995) 67–74. B. HURTGEN, C. STILLER, FAST HIERARCHICAL CODEBOOK SEARCH FOR FRACTAL CODING STILL IMAGES, SPIE VISUAL C OMMUN. PACS MED. APPL. (1993) 397–408. C. HUFNAGL, A. UHL, ALGORITHMS FOR FRACTAL IMAGE COMPRESSION ON MASSIVELY PARALLEL SIMD ARRAYS, REALTIME IMAGING 6 (2000) 267–281. D. VIDYA, R. PARTHASARATHY, T.C. B INA, N.G. SWAROOPA, ARCHITECTURE FOR FRACTAL IMAGE COMPRESSION, J. SYST. ARCHIT. 46 (2000) 1275–1291. Y. FISHER, FRACTAL IMAGE COMPRESSION, SIGGRAPH’92 COURSE NOTES 12 (1992) 7.1–7.19.

Goldhill

0.5

Camera man

0.5

242

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

**(IJCSIS) International Journal of Computer Science and Information Security, Vol. 8, No. 2, 2010
**

Dr.R.S.D.Wahidabanu obtained her B.E. degree in A.Muruganandham obtained her B.E. degree in 1993 in from Madras University and M.E. degree in 2000 from Bharathithasan University and doing Ph.D in Anna University, Coimbatorei. His areas of interest are Image processing, He is a life member of ISTE and member of IEEE. He is currently working as a Asst. Professor and Electronics and Communication Engineering Sona College of Technology, Salem-05, Tamilnadu, India and has a teaching experience of 15 years. Authored national and international conferences and journals. 1981 and M.E. degree in 1984 from Madras University and Ph.D in 1998 from Anna University, Chennai. Her areas of interest are Pattern

Recognition, Application of ANN for Image Processing, Network Security, Knowledge

Management and Grid Computing. She is a life member of IE, ISTE, SSI, CSI India and ISOC and IAENG. She is currently working as a Professor and Head of Electronics and Communication Engineering, Government College of Engineering, Salem, Tamilnadu, India and has a teaching experience of 27 years. Authored and coauthored 180 research journals in national and international conferences and journals.

243

http://sites.google.com/site/ijcsis/ ISSN 1947-5500

- Journal of Computer Science IJCSIS March 2016 Part II
- Journal of Computer Science IJCSIS March 2016 Part I
- Journal of Computer Science IJCSIS April 2016 Part II
- Journal of Computer Science IJCSIS April 2016 Part I
- Journal of Computer Science IJCSIS February 2016
- Journal of Computer Science IJCSIS Special Issue February 2016
- Journal of Computer Science IJCSIS January 2016
- Journal of Computer Science IJCSIS December 2015
- Journal of Computer Science IJCSIS November 2015
- Journal of Computer Science IJCSIS October 2015
- Journal of Computer Science IJCSIS June 2015
- Journal of Computer Science IJCSIS July 2015
- International Journal of Computer Science IJCSIS September 2015
- Journal of Computer Science IJCSIS August 2015
- Journal of Computer Science IJCSIS April 2015
- Journal of Computer Science IJCSIS March 2015
- Fraudulent Electronic Transaction Detection Using Dynamic KDA Model
- Embedded Mobile Agent (EMA) for Distributed Information Retrieval
- A Survey
- Security Architecture with NAC using Crescent University as Case study
- An Analysis of Various Algorithms For Text Spam Classification and Clustering Using RapidMiner and Weka
- Unweighted Class Specific Soft Voting based ensemble of Extreme Learning Machine and its variant
- An Efficient Model to Automatically Find Index in Databases
- Base Station Radiation’s Optimization using Two Phase Shifting Dipoles
- Low Footprint Hybrid Finite Field Multiplier for Embedded Cryptography

Sign up to vote on this title

UsefulNot usefulThe Fractal image compression encodes image at low bitrate with acceptable image quality, but time taken for encoding is large. In this paper we proposed a fast fractal encoding using particle swar...

The Fractal image compression encodes image at low bitrate with acceptable image quality, but time taken for encoding is large. In this paper we proposed a fast fractal encoding using particle swarm optimization (PSO). Here optimization technique is used to optimize MSE between range block and domain block. PSO technique speedup the fractal encoder and preserve the image quality.

- www.ijerd.com
- Fractal Image Compression Using Quatree Partition
- Fractal Image Compression
- prl_05
- 13 Image Compression
- Biogeography-Based Optimization Technique for Block-Based Motion Estimation in Video Coding_2
- Data Compression and Huffman Algorithm
- Compression Algorithms for Medical Images
- Compress.4up
- STILL IMAGE COMPRESSION
- High Data Rate Transmission in Cdma by Image Compression Using Mean Square Error Method
- JPEG Compression Stegenography
- JPEG Image Compression and Decompression by Huffman Coding
- Lt 3620172020
- Reversible Data Hiding VIA Optimal Code for Image
- Data Compression Wp
- DATA COMPRESSION AND HUFFMAN ALGORITHM.ppt
- 20Compression
- Bch Codeing Techniques
- imagecmprs
- Golomb Code
- VOICE PREDICTION AND SPEECH COMPRESSIONUSING ADSP2105
- Application of Steganography in Medical Images
- EAC Mp3 Vorbis Setup Guide
- What Does MP3 Stand For?
- ppctual_coder.pdf
- 1
- Towards block-based compression of genomic data with random access functionality
- The Emerging JBIG2 Standard
- Compression
- Effective MSE optimization in fractal image compression