DPCM HYBRID CODING FOR INTERLACED IMAGE

121

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

DCT/DPCM HYBRID CODING FOR INTERLACED IMAGE COMPRESSION Dr. Saied Obied Abdul – Amir Assistant Prof Communication &Computer Dept. Al Mansour College

Khamies Khalaf Hasan Assistant Lecturer Electrical Eng. Dept. University of Tikrit

ABSTRACT By the nature of images, picture elements in local regions are highly correlated with one another. In such cases, image compression techniques are introduced to reduce the amount of data is needed to represent the same information, either exactly or approximately. In this work DCT/DPCM hybrid approach have been designed and implemented for interlaced images. The image signal was first transformed row-wise using discrete cosine transform (DCT) and a differential pulse code modulation (DPCM) scheme then was used column-wise to get difference signal. For still images the same 3-bit quantizer was employed which makes quantization process easier. For interlaced images 3-bit quantizer was used for the odd field and 2-bit quantizer for even field, since the difference signal of the even field was very small. A compression ratios of about 13:1 was obtained for interlaced image. Objective measurements showed a high peak to peak signal to noise ratio without noticeable impairment.

KEY WORDS: Interlaced, Hybrid image compression, DCT/DPCM, Discarding.

List of Abbreviations Abbreviation

Meaning

Abbreviation

Meaning

BR

Bit Rate Compression Ratio

1-D

One Dimensional

CR

2-D

Two Dimensional

DCT

Discrete Cosine Transform

bpp

Bit per pixel

DFT

Discrete Fourier Transform

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

Abbreviation DPCM

Meaning

122

Differential Pulse Code Modulation

List of symbols

HVS

Human Visual System

Symbol

Description

IDCT

Inverse Discrete Cosine

dq

Quantized difference

I (r )

The one dimensional function

I ( r , c)

The two dimensional original image

. I(r,c)

The decompressed image.

Transform MSE

Mean Squared Error

PSNR

Peak to peak Signal to Noise Ratio

PCM

Pulse Code Modulation

RLC

Run Length Coding

SQ

Scalar Quantizer

L Number of gray levels

SNRRMS HVS IDCT

123

The root-mean-square signalto-noise ratio. Human Visual System Inverse Discrete Cosine Transform

MSE

Mean Squared Error

PSNR

Peak to peak Signal to Noise Ratio

PCM

Pulse Code Modulation

RLC

Run Length Coding

SQ

SNRRMS

Scalar Quantizer The root-mean-square signalto-noise ratio.

ri

Quantizer reconstruction levels.

S

Actual (present) sample value.

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132) 123

INTRODUCTION Compression of digital images

full

frame.

Interlacing

works

by

has been a topic of research for many

scanning every odd line on the screen,

years

followed by every even line in the

and

compression

a

number

standards

of

image

have

been

second scan. [5, 6]

created for different applications. A

The image quality is measured

digital mage can be considered as two-

either subjectively or objectively. The

dimensional array of samples I(r, c). [1, 2]

parameters that decide if a picture to be

Due to the errors introduced by

considered is of high quality will differ

sampling and quantization of the actual

when either of them is used.

scene, it is only an approximation of an

The subjective visual quality

actual scene could be obtained. Number

measurement plays an important role in

of levels of intensity determines the

visual communications. It is natural that

Precision are expressed as the number

human viewers should judge the visual

of bits/sample. [3, 4]

quality of reconstructed images or video

It

is

well

known

that

an

analogue television picture is built up of

frames since they are the ultimate receivers of the data. [7]

lines. These pictures or frames are

Objective quality measurements

updated with a certain frequency (25 Hz

are

or 30 Hz). At such a low frame rate the

instrumentation. All types of objective

flicker in the TV picture is annoying. A

video or image quality assessments are

way to solve this problem would be to

done by measuring the distortion as the

increase the number of frames per

difference (error) between the original

second, to say 50 Hz; a method to

and the reconstructed image by a

increase the video update frequency and

predefined function.:

to lower the bandwidth usage is to use interlaced pictures. Interlacing doubles

conducted

by using

. error(r,c)  I(r,c) - I(r,c)

electrical

(1)

the frame update frequency, the odd and the even field of each frame is transmitted consecutively. Each field contains only half the information of a

Where: I(r, c) = the original image,

. and I(r,c) = the decompressed image

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

124

The root-mean-square error is defined by square root of the error

The bit rate and compression ratio are simply related by: [9]

squared divided the total number of

BR  (BPP for origional image) CR (5)

pixel in the image ( N 2 ): Another related metric, the Peak Signal-to-Noise Ratio (PSNR) metric, is defined as:

Compression algorithms can be divided

into

lossless

and

lossy

algorithms depending on the type of PSNR dB  10 log10

………(2)

(L  1) 2 1 N2

N 1 N 1 .

   I(r, c) - I(r, c)  r 0 c 0 

coding used which will be determined

2

by the particular needs of the user. Lossless compression of images

Where L = the number of gray levels

involves a completely reversible scheme

(e.g., for 8 bit L = 256) These objective measures are

by which the original data can be

often used in research because they are

reconstructed exactly; it can achieve a

easy

compression ratio of up to 3:1. Run-

to

generate

and

seemingly

Length coding and Huffman coding are

unbiased. [8] The most obvious measure of compression

efficiency

is

the

some of the lossless compression methods used in image compression. [10] In

Compression Ratio ( CR ) metric which

compression

is often employed and is defined as:

order

to

ratios

achieve with

high

complex

images, lossy compression methods are

CR 

original image compressed image (3)

required. Lossy compression provides trade off between image quality and degree of compression.

is

In fact many lossy compression

instead quantified by stating the Bit

techniques are capable of compressing

Rate BR achieved by compression in

images with 10:1 to 20:1 and still retain

bpp (bits per pixel) and is defined by:

high quality visual information.. The

Sometimes

compression

lossy

original image size in bit BR  No of pixels of the image (4)

methods

coding, transform

include

predictive

coding and the

combination of both which is called hybrid coding. [11]

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132) 125

EXPERIMENTAL WORK The hybrid system consists of

blocks of the same length. Each block is

five main stages, see Fig. (1), firstly the

considered as a row vector, considering

encoding process of the input still image

the odd field represented by every odd

into interlaced image by dividing the

line in the image is the first field to be

original image into two fields. Then

coded.

each line in the field is divided into

1D DCT is applied to transform

coefficients

useful

(previous

spatial domain to frequency domain

coefficients, see Fig. (2), the third order

with

prediction is not possible.

same

dimension

to

be

represented in a more compact form.

energy, where the high frequency

vertical

odd

The predicted value of T4 is then

The few low frequency coefficients represent the DCT packing of the image

next

prediction

each row vector in the odd field from

the

and

for

given by: 





T4  a7 T  a8 T2

(6)

coefficient can be discarded with little loss

in

coefficients

energy; are

the

discarded

replaced

by

zero

.

Where:

.

T and T2

is

the

reconstructed coefficient of T and T2

coefficients. See Fig. (2). As mentioned previously that

respectively.

coefficients

same

prediction weighting coefficients for the

highly

even field. The prediction difference is

correlated, a one dimensional first order

then fed to the quantizer to obtain the

DPCM is used for encoding the odd

quantized difference which adds to the

field. The predictor uses the vertical

predicted value to get the reconstructed

coefficient value for the prediction

coefficient. Each field quantized in

according to the equation (6).

different quantizer, (2bits) quantizer is

DCT

horizontal

with

frequency

the are

But, for even field second order predictor is used because the current coefficient

has

two

neighboring

a 7 and

a8

are

the

used with the even field, while as for odd field (3bits) quantizer is used.

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

The same DPCM system is used

Step 8: End. 126

to decode each field coefficients, and then the DPCM decoding coefficients is applied to 1D-IDCT to reconstruct the

SIMULATION RESULTS

original interlaced image data. The

To evaluate the hybrid system

hybrid system (DCT/DPCM) for the

performance with the applied Saturn

interlaced image is illustrated in the

image of size (256×256) with 256 gray

algorithm below.

levels. Two major measurements have

Algorithm 1:

been employed, namely the objective

Input: The original interlaced image

and subjective tests.

Output: The Compressed interlaced image

In this system each line is divided into blocks. Each block is

Step 1:

Selection of the original

considered as a row vector of 128

interlaced image.

pixels. 1D DCT is applied to convert

Step 2: Divide each field of the

each vector of pixels into a vector of

interlaced image into row vectors.

128 transformed coefficients. 1D DCT

Step 3: Apply One Dimension-Discrete

packing the energy into few numbers of

Cosine Transform (1D DCT) to each

transformed coefficients associated with

field row wise considering odd field

low frequencies, the high frequency

first.

coefficients are discarded and replaced

Step 4: Discarding the DCT coefficients

by zero coefficients. The discarding

that represent high frequencies.

process provides a tradeoff between CR

Step 5: Apply DPCM coding to DCT

and PSNR.

coefficients using first order prediction

Fig.3 ( b, c, d and e) show

for odd field , Second order prediction

the Saturn reconstructed images

for even field.

using

(64,

51,

low

frequency

39

and

32)

Step 6: DPCM reconstruction of the transformed coefficients.

coefficients

Step 7: Inverse DCT: Apply the inverse

respectively. They show a higher

1D

reconstructed

subjective quality. Comparing the

coefficients obtained from step 6 to get

reconstructed images with the

the reconstructed image.

original image, no noise in smooth

DCT

to

the

128

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

127

area and no edge degradation are

Reduction”

evident.

California, PhDThesis,2000.

(Table 1) list the objective test based on CR and PSNR. Generally when the bit rate decreases the PSNR

University of

Southern

Website: http://sipi.usc.edu/~ortega/PhDTheses/C hristosChrysafis.pdf

will decreases and the quality will be corrupted by smearing in the edge

[2]- Salomon, D., “Data Compression ”, Springer-Verlag, second edition, 2000.

CONCLUSIONS From the results obtained, one can conclude the following points:

[3]- Gonzalez, R. C., and Woods, R. E., “Digital Image Processing”, Pearson Education Asia Pte Ltd., 2000.

1-The actual efficiency of the compression system depends to some extent on the original image quality.

[4]- Pasi, F “Image Compression”, “University of Joensuu”, lecture notes, 2002.

2-The discarding process of high frequency DCT coefficients provides a trade off between CR and image

[5]- Kaxe, B., “Synchronization of MPEG-2 Based Digital TV Services Over IP Networks”, Master’s thesis,

quality.

2000.Website:http://vvv.it.kth.se/docs/R 3-Using DCT by segmenting each line

eports/

into vectors may result in blocking artifacts. These artifacts are perceptually annoying and become prominent in the reconstructed images at very low bit-

[6]- Motta, G., “ Optimization Methods For

Data

Compression”,

Brandeis/University,PhD/Thesis,2002. Website:www.cs.brandeis.edu/~gim/Pa

rates.

pers/OptimizationMethodsForDataCom REFERENCES

pression.pdf

[1]-Chrysafis, C., “Wavelet Image

[7]- Lislevand, H., “The Connection

Compression

Rate

Between

Optimization

and

Distortion Complexity

Compression

Needs

,

Bandwidth and Quality for Multimedia Transfer in 3G System “ , Agder

128

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

University College “ , Master’s thesis,

[10]- Subhi, A., “Interframe Coding for

2004 .

TV Signals”, Master’s thesis, College of Engineering,

[8]-Scott E. Umbaugh, “Computer

Al-Nahrain

University,

2000.

Vision and Image Processing”, Prentice Hall PTR, 1998.

[11]

Ali,

Compression [9]-Adamson,C.,”Lossless Compression of Magnetic Resonance Imaging Data”, Monash University, Master’s thesis, 2002.

A.

H., Using

”Image

Data

Adaptive

Methods”. University of Technology, Master’sthesis,2000

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

129

Table (1): Objective Results for interlaced Saturn image.

Compression ratio

Bit per pixel (bpp)

PSNR (db)

64 column

6.4

1.25

36.2834

51 column

8.0314

0.9961

35.6303

10.5026

0.7617

34.6501

12.8

0.625

33.6895

Number of selected column /vector

39 column

32 column

Original egami

Reconstructed image

Interlaced image coding

DCT Transformation

Inverse DCT Transformation

DPCM Coding

DPCM Reconstruction

Fig. (1) Block diagram of the interlaced hybrid (DCT/DPCM) system.

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

130 130

Increasing frequency odd

S3 S2

odd

deven

S5 S4

deven

T5 T4

odd

S1 So

odd

T1 To

deven

even d

odd

odd

deven

deven

odd

odd

d (a) The dots ( ) represent the pixels values (S)

T3 T2

d dots ( ) represent the coefficients values (T) The

discarded coefficients Zero’s odd deven (c)

odd deven odd deven odd d

(b)

Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132)

131

Fig. (2) Procedures applied to each (row vector) of the interlaced image

(a) The original interlaced image. (b) 1D Transformed image. (c) Transformed coefficients after discarded high frequency coefficients.

(a)

(b)

(c)

(d)

(e) Fig. (3) Interlaced hybrid DCT/DPCM of Saturn image.

‫)‪Tikrit Journal of Eng. Sciences/Vol.16/No.1/March 2009, (121-132‬‬

‫‪(a) Original image with (8bpp).‬‬ ‫‪132‬‬

‫‪(b) Reconstructed image with 64 columns/block (1.25 bpp).‬‬ ‫‪(c) Reconstructed image with 51 columns/block (0.9961bpp).‬‬ ‫الترميز ألهجيني‪ DCT/DPCM‬في ضغط الصور المتشابكة‬ ‫خميس خمف حسن‬

‫د‪.‬سعيد عبيد عبد األمير‬

‫مدرس مساعد‬

‫أستاذ مساعد‬

‫قسم هندسة االتصاالت والحاسبات‬

‫قسم الهندسة الكهربائية‬

‫المنصور‬

‫كمية‬

‫جامعة تكريت‬

‫الخالصة‬ ‫من خالل بيعةالا صورال تك ن ال ن ر ارالت صورال تن طال م الابم ماالططن منتصيبالا ي ال‬ ‫هذه صواالاتهك رميعالاه ضالطب صورال ت نخالنخطت ون يعال‬ ‫ناما أ ن تعيعاف نت نرمعت ن لعذ ظات هجع‬ ‫ياخالنخطصت ماال‬

‫صوجعالالم صوم لرال‬

‫معالا صومةبعالاه صومبي يالا ونم عال‬

‫يعالت مال يةضالاا صوالية‬

‫ف طال‬

‫لالع صومةي مالاه ومالا يرال تن‬

‫وضطب صور ت صومن اي اف عنت نا عال و الاتن صورال تن يانجالاه صورال‬

‫الالت عخالنخطت صونم عال صوتمالال صونلاضالالي صوت مال يانجالالاه صوةمال ط ويارال‬

‫ريالة و الالاتن‬

‫صولتم فياو خيا وير ت صوخا ا نت صخنخطصت م مت رالططل أمالا ويرال ت صومن الاي ا ط الط نالت صخالنخطصت لالع صوم مالت صوةالططل‬

‫ويمجالالا صولالالتطل‬

‫م مالالت رالالططل ص ال ويمجالالا صوا ج ال تن و الالاتن صولالالتم ويمجالالا صوا ج ال‬

‫ا الاله رالالطعتن جالالطصك أربالاله‬

‫صو عاخاه صوةميعا عما راوعا و خيا صإل اتن ووة صوض ضاء)‪ (PSNR‬يط ن أل ن ه ميا ظف‬ ‫الكممات ألدالة‪ :‬صومن اي اك ضطب صور ت صواجعنك صونتمعا أواجع‬

‫‪DCT/DPCM‬ك بتح صولائ‬

‫ف‬