219-523-1-PB

Venkat Narayana Pasula* et al. / (IJITR) INTERNATIONAL JOURNAL OF INNOVATIVE TECHNOLOGY AND RESEARCH Volume No. 1, Issue No. 5, August -September 2013, 509 - 513.

ISSN 2320 5547 @ 2013 http://www.ijitr.com All rights Reserved. Page | 509

EFFICIENT IMAGE COMPRESSION USING SPIHTAND HUFFMAN CODING BASED ON DWT

THROUGH TRANSMITTING OFDM CHANNELVENKAT NARAYANA PASULA

M. Tech (DECS), Department of ECE, GurunanakInstitutions Technical Campus

C HEMASUNDHARA RAO,Professor, Department of ECE, Gurunanak

Institutions Technical Campus

ABSTRACT:- In this paper we tend to gift associate energy saving approach to transmission of separatemoving ridge transformation primarily based compressed image frames over the OFDM channels. Supportedchannel state data at the Transmitter, the descriptions so as to the method of excellent channels utilized inHuffman and SPHIT encryption. Analysis in analysis of the system in terms of chance of error is disbursed ina very subtle wireless optical channel. As per planned system shows promising results for a high speed opticalwireless channel and that we demonstrate the quality of our planned theme in terms of system energy savingwhile not compromising the received quality in terms of peak signal-noise quantitative relation. Despitevictimization a lot of range of carriers rather than set of carriers is usable for made knowledge transmissionand allowing the re-transmission of lost packets. Transmitted image and its received versions at totallydifferent PSNRs then it'll improve additionally. The enhancements which will be accomplished in variedperformance parameters in a very data communication system victimization moving ridge transformationmethod. Parameters like MSE PSNR COMPRESSION RATIO power spectral density and Real, notionalelements of the OFDM signal Verification of Efficiency analysis. Huffman and SPIHT decoding of embeddedencoder. In this technique reduce the number of encoding bits and reduce the system power consumptions.

Keywords:- Image. DWT. Huffman, SPIHT, OFDM

I. INTRODUCTION

OFDM may be a multi-carrier modulation themehaving wonderful performance that allowsoverlapping in frequency domain. In OFDM,individual sub channels square measure plagued byflat weakening, thus for a quantity of some time,condition of the sub channels could also be sensible,or they will be deeply pale. The packets that squaremeasure transmitted through these pale sub channelssquare measure extraordinarily at risk of be lost at thereceiver as a result of non-acceptable errors. OFDMsystem provides an opportunity to use the variety infrequency domain by providing variety ofsubcarriers, which could work as multiple channelsfor applications having multiple bit streams. Onestraightforward overlade en transmission answer toupgrade existing OFDM based mostly broadcastmulticast traffic channel. With this theme, heritagemobiles will seamlessly operate within the upgradednetwork while not further modification. Themanagement over head signal half is same. The pilothalf is reused. solely the traffic channel half isupgraded. The new traffic channel half is layer-modulated and transmitted with a further pre-codedOFDM modulated sweetening layer, wherever the s-symbols square measure pre-coded with Walsh-Handmaid matrix before OFDM. In associateadditive white Gaussian channel, this theme has thesuperposition pre-coding (SPC) gain since it

primarily is associate implementation of SPC.However, the interference from the sweetening layeris irregular because of further Walsh-Handmaidspreading. Within the weakening channel, furthermulti-layer diversity gain is doable, since the bottomlayer and also the sweetening layer square measure inoperation in several signal areas.

II. IMAGE

Digital image is printed as a two dimensional operatef(x, y), where x and y area unit special (plane)coordinates, and thus the amplitude of f at any striveof coordinates (x, y) is termed intensity or greylevel of the image at that point. The digital imageconsists of a finite vary of parts, each of that has aparticular location and value. {The parts the weather}area unit declared as image parts, image parts, pixels,and pixels. pel is that the term most usually used.Digital compression addresses the matter of reducingthe number of data required to represent a digitalimage. The underlying basis of the reductiontechnique is removal of redundant data. a picturecompressed employing a lossless technique is sure tobe just like the initial image once uncompressed.

III. COMPREESION PROCESS

Lossy schemes, on the opposite hand, throw uselessknowledge away throughout coding. This is, in fact,however lossy schemes manage to get superiorcompression ratios over most lossless schemes. JPEG



was designed specifically to discard info that thehuman eye cannot simply see. Slight changes in colordon't seem to be perceived well by the human eye,whereas slight changes in intensity (light and dark)area unit. so JPEG's lossy coding tends to be a lot ofstinting with the gray-scale a part of a picture and tobe a lot of dizzy with the colour.JPEG was designedto compress color or gray-scale continuous-tonepictures of real-world subjects: pictures, video stills,or any complicated graphics that correspond naturalsubjects. Animations, ray tracing, line art, black-and-white documents, and typical vector graphics do notcompress fine below JPEG and should not beexpected to. And, though JPEG is currentlyaccustomed give motion video compression, thequality makes no special provision for such ANapplication.

Fig:1 Block Diagram:DWT +Huffman and SPIHT using Image Transmission in OFDM Channel

The fact that JPEG is lossy and works solely on a getstyle of image knowledge may cause you to raise,"Why hassle to use it?" It depends upon your desires.JPEG is a wonderful thanks to store 24-bitphotographic pictures, like those employed inimaging and transmission applications. JPEG 24-bit(16 million color) pictures ar superior in look to 8-bit(256 color) pictures on a VGA show and ar at theirmost spectacular once exploitation 24-bit showhardware (which is currently quite inexpensive).Theamount of compression achieved depends upon thecontent of the image knowledge. A typicalphotographic-quality image could also be compressedfrom 20:1 to 25:1 while not experiencing anynoticeable degradation in quality. Highercompression ratios can end in image files that dissentperceptibly from the first image however still haveAssociate in Nursing overall smart image quality.And achieving a 20:1 or higher compressionmagnitude relation in several cases not solely savesdisc space, however additionally reduces coordinated

universal time across knowledge networks and phonelines.

DWT Process in Image

Fig: 2 DWT process in input image

In the case of lossy compression, progressive or not,JPEG2000 with not reversible Daubechies 9,7 filtersis better than all the other standards in terms of PSNRon the average MSE, whatever the bit rate is: byincreasing the compression ratio, the quality ofJPEG2000 increases in comparison with the otherones. In terms of visual quality, JPEG2000 wascompared with JPEG only. The result shows that,visual quality being equal, JPEG needs a larger bitrate (from 13% to 112% depending on the number ofbits) and the differences increase if the compressionincreases.

Huffman Process

Huffman secret writing uses a selected methodologyfor selecting the illustration for every image, leadingto a prefix code (sometimes known as "prefix-freecodes", that is, the bit string representing someexplicit image is rarely a prefix of the bit stringrepresenting the other symbol) that expresses theforemost common supply symbols victimizationshorter strings of bits than are used for fewercommon supply symbols. Huffman was able to stylethe foremost economical compression methodologyof this type: no different mapping of individualsupply images to distinctive strings of bits canmanufacture a smaller average output size once theparticular symbol frequencies believe thoseaccustomed produce the code. it's a appliedmathematics secret writing. it's allotted all importantcoefficients with constant area.

For example: 212 coefficient = 1 byte.



In encoder, produce binary tree nodes with characterand frequency of every character then the Placenodes in an exceedingly priority queue. Importantconstant and convert to bit stream. In decoder, oncereceiver has at bay it scans incoming bit stream andto saves the ton of bits. Here a concrete example toinvestigate the output binary stream of SPIHTcryptography. The subsequent is three level riffledecomposition coefficients of SPIHT encoding;

SPIHT Process

It is a picture compression formula supported 3concepts :(1) Partial ordering of the remodeled imageparts by magnitude and transmission of this orderingdata.(2) Ordered bit plane transmission.(3)Application of similarity between coefficients fromcompletely different ripple levels that describeidentical origin

In the SPIHT formula, compression is complete in 2ways that. initial of all, as a result of the remodeledimage parts ar part ordered by magnitude, the leading0 bits and also the initial 1 of any constant don'tought to be transmitted, since they'll be derived fromthe ordering data. Secondly, the SPIHT-algorithmproduces Associate in embedded bit stream. InAssociate in embedded bit stream the secret writingwill be stopped at any time and still the image will bedecoded and reconstructed. Moreover, if theremodeled image is absolutely encoded and decoded,the reconstruction of the image is lossless.

OFDM Process

Orthogonal frequency division multiplexing may be adigital modulation technique. In digital modulationthe signal is split into several slender band channels.And that they have completely different frequencies.Orthogonal frequency division multiplexing is specialvariety of multi carrier modulation technique.

Fig: Block fading channels in OFDM system in Frequency Sub-Carrier

MIMO wireless system consists of N transmitantennas and M receive antennas. However, not likephased array systems wherever one info stream, sayx(t), is transmitted on all transmitters and so receivedat the receiver antennas, MIMO systems transmitcompletely different info streams, say x(t), y(t), z(t),

on every transmit antenna. These square measurefreelance info streams being sent at the same timeand within the same waveband. Initially look, onemay say that the transmitted signals interfere witheach other. In reality, however, the signal inward atevery receiver antenna is a linear combination of theN transmitted signals.

IV. RESULT ANALYSIS

BER Process

From the experimental results, we can see that valuesof L are less than 3, so we can achieve thecompression effect. For each image in the same ratealways the probability of each symbol appear flat,and only small fluctuations, so saving the number ofbits are also pretty much the same thing.

Fig: 5(a) Compression Process in OFDM Channel

With the rate increase word code length in average(L)will be an increasing trend, but after the rate greaterthan 0.3bpp the trending will be become very slow ,and more value of rate more bits will be save.

As an example, assume this transmitted bit sequence:

0 1 1 0 0 0 1 0 1 1,

and the following received bit sequence:

0 0 1 0 1 0 1 0 0 1,

Fig: 5(b) OFDM Transmission Signal

The number of bit errors (the underlined bits) is inthis case 3. The BER is 3 incorrect bits divided by 10transferred bits, resulting in a BER of 0.3 or 30%.



Mean Square Error (MSE):-

It is defined as the square of error between originalimage and the compress image. The distortion in theimage can be measured using MSE.

MSE= [A(i,j )- B(i,j)]^2

N X NHere, A(i.j)= Original image.

B(i.j)= Compress image.

N X N=row and column of image intensity of pixelvales (255 255) image size.

Peak Signal to Noise Ratio

It is the ratio of the maximum signal to noise in thecompress image

PSNR=20log10 {(255X255) / (MSE)}

Efficiency Calculation

It is defined as the ratio between compressed file sizeand input file size.

Eff = (Comp_filesize / input_filesize)*100;

Comparison Table: Compression Technique

PSNR

CompRatio

MSE

DWT+HUFFMAN 27.65 0.4410

111.69

DWT+SPIHT 44.83 1.1395

2.1417

DWT+HUFFMAN+SPIHT

48.13 1.3472

0.9995

V. CONCLUSION:-

Proposing an easy and effective technique combinedwith Huffman coding for additional compressionduring this paper that saves plenty of Log bits withinthe image information transmission. Theres terriblybig selection of sensible price for these days thatcontains a sizable amount of image datas to betransmitted. We have a tendency to propose a energysaving approach, wherever the compressedcoefficients ar organized in down order of priorityand mapped over the channels beginning with thegreat ones. The coefficients with lower importancelevel, that ar probably mapped over the dangerouschannels, ar discarded at the transmitter to avoidwasting power while not vital loss of receptionquality. Our analytic observations on receptionquality and energy saving performance ar valid byintensive MATLAB simulations. In the future, thiswork will more be optimized to hold out all thephases on Verilog HDL. The Verilog HDL code is

often optimized to attain the required leads to lessertime and fewer storage parts. Also, the newer FPGAfamilies being introduced within the market providehigher performance and additional storage parts toreinforce performance and yield higher results.

REFERENCES

[1] Rafael C. GONZALEZ Richard E. WOODS.Digital image processing: second male erectiledysfunction [M]. National capital businessfirm of industry 2002

[2] Brandy ANTONINI Michel BARLAUD PierreMATHIEU et al. Image committal to writingvictimization rippling remodel [J]. IEEETrans. Image process 1992 1(2) 205-220

[3] Cheng Li-chi, Wang Hong-xia, Luo Yong.Rippling theory and applications. Beijing:Science Press, 2004(Chinese)

[4] J. M. SHAPIRO. Embedded image committalto writing victimization zero trees of waveletscoefficients [J]. IEEE Trans. Signal process1993 41(12) 3445-346 a pair of.

[5] Rafael C. GONZALEZ Richard E. WOODS.Digital image processing: second maleerecticle dysfunction [M]. National capitalbusiness firm of industry 2002.

[6] M. Banerjee and M. K. Kundu, Edgeprimarily {based} options for content basedimage retrieval, Pattern Recognition., vol. 36,no. 11, pp. 26492661, November 2003. [8] Y.S. Chan, P. C. Cosman, and L. B. Milstein, Across-layer diversity technique for multi-carrier OFDM multimedia system networks,IEEE Trans. Image Proc., vol. 15, no. 4, pp.833847, Apr. 2006.

[7] R. J. McEliece and W. E. Stark, Channelswith block interference,IEEE Trans. info.Theory, vol. 30, no. 1, pp. 4453, Jan. 1984.

[8] Richard Van inheritable and Ramjee Prasad,"OFDM for wireless multimedia systemcommunications," Arech House Beantown,London, 2000.

[9] J. M. Kahn, W. J. Krause, and J. B. Carruthers,"Experimental characterization of non directedindoor infrared channels," IEEE Trans.Commun., vol.43, pp.1613-1623, 1995.

[10] Weinstein. S. B and Ebert P. M, "Datatransmission by frequency divisionmultiplexing victimisation the separate Fourierremodel," IEEE transactions on



Communications, vol-com-19, pp. 628-634,1971.

[11] R. McEliece and W. E. Stark, Channels withblock interference, IEEE Trans. Inform.Theory, vol. IT-30, pp. 4453, Jan. 1984.

BIBILOGRAPHY

Venkat Narayana Pasula did his B.Tech in ECE fromPrinceton College of Engineering& Technology (JNTUH) andperusing the M.Tech in DECSfrom Gurunanak InstitutionsTechnical Campus, Hyderabad in2013. His areas of interest inresearch are Digital ImageProcessing & Signal Processing

Mr.C.Hemasundara Rao did his M.Tech from JNTUHyderabad and Perusing Ph.Dfrom Kakinada University. He isworking as a Professor in thedepartment of Electronics &Communication Engineering,Gurunanak Institutions ofTechnical Campus Hyderabad,

India. His areas of interest in research are ImageProcessing, VLSI & Signal Processing.

Documents

219-523-1-PB