International Journal of Graphics JOURNAL and Multimedia (IJGM), ISSN 0976 AND 6448(Print), INTERNATIONAL OF GRAPHICS ISSN 0976

6 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME

MULTIMEDIA (IJGM)

ISSN 0976 - 6448 (Print) ISSN 0976 -6456 (Online) Volume 4, Issue 1, January - April 2013, pp. 01-08 IAEME: www.iaeme.com/ijgm.asp Journal Impact Factor (2013): 4.1089 (Calculated by GISI) www.jifactor.com

IJGM
IAEME

A NOVEL TECHNIQUE IN SPIHT FOR MEDICAL IMAGE COMPRESSION

Shipra Gupta1, Chirag Sharma2
1 2

(Computer Science and Engg , Lovely Professional University, Hoshiarpur, India) (Computer Science and Engg, Lovely Professional University, Jalandhar, India)

ABSTRACT Medical science grows very fast and each hospital needs to store high volume of data about patients and in this field the images produce by the modality is in the form of large file, in order to get the opinion from other doctors images are send to other place using electronic media. Compression of images needs to be apply as the size of image is very large to send, but with compression there is loss of information in the image. In order to minimize the loss and to increase the quality of image requires compression is to be done, multi wavelet transformation technology plays a vital role. So, in this paper we consider that multi wavelet with Region of Interest (ROI) on the selecting portion will not only give the quality but also reduce the loss of information from image. And we are going to implement the multi wavelet transformation with Modified Fast Haar Wavelet Transform (MFHWT) in Set Partitioning in Hierarchical Trees algorithm (SPIHT). Keywords: Medical Image, MFHWT, Multi wavelet, ROI, SPIHT. I. INTRODUCTION

Image compression is the process of encoding information using fewer bits. Compression is useful because it helps to reduce the consumption of expensive resources, such as hard disk space or transmission bandwidth. It also reduces the time required for images to be sent over the Internet or download from web pages. It also helps in accelerating transmission speed [1]. Data compression methods are usually classified as either lossless or lossy methods [1] [8].

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME 1. Wavelet Transform

When the signal in time for its frequency content is analyzed then in that wavelet functions are used. Multi resolution hierarchical characteristics are provided by wavelet based compression. Hence image can be compressed at different levels of resolution. It can be sequentially processed from low resolution to high resolution [1] [8]. It has excellent energy compaction property which suitable for exploiting redundancy in an image to achieve compression [2] [8]. Wavelets are localized in the both time and frequency domains. Hence it is easy to capture local features in a signal [1] [8]. A newer alternative to the wavelet transform is the multi wavelet transform. Multi wavelets are similar to wavelets but have some important differences. In particular, whereas wavelets have an associated scaling function and wavelet function, multi wavelets have two or more scaling and wavelet functions [3].

Fig. 1 (a) Wavelet (b) multi wavelets [3] [8]

2.

Haar Transform

The Haar wavelet transformation is a simple form of compression involved in averaging and differencing term, sorting detail coefficients; eliminate data and reconstructing the matrix such that the resulting matrix is similar to initial matrix [4]. 3. Modified Fast Haar Wavelet Transform (MFHWT)

MFHWT can be done by just taking (w+x+y+z)/4 instead of (x+y)/2 for approximation and (w+x-y-z)/4 instead of (x-y)/2 for differencing process. 4 nodes are considered at a time [1]. Also, it is used to reduce the memory requirements and the amount of inefficient movement of Haar coefficients [5]. Thus MFHWT reduce the calculation work of Haar transform. 4. SPIHT

SPIHT algorithm is one of the powerful algorithm for the compression. After wavelet transform SPIHT algorithm is used to encode the coefficients of wavelet. In SPIHT sorting is done by comparing two elements at a time and results in yes/no states. In this sorting pass coefficients are categorizes into 3 lists [8]:
2

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME LIS List of Insignificant sets are the set of coefficients having magnitude smaller than the threshold. LIP List of Insignificant Pixels are the coefficients having magnitude smaller than the threshold. LSP List of significant pixels are the pixels those magnitude is larger than that of threshold. In this pass, only bits related to the LSP entries and binary outcomes of the magnitude tests are transmitted to the decoder. In implementation, we grouped together the entries in the LIP and LIS which have the same parent into an entry element. For each entry element in LIP, we estimated a pattern in both encoder and decoder to describe the significance status of each entry in the current sorting pass. If the result of the significance test of the entry item is the same as the specified pattern, we can use one bit to represent the status of the whole entry atom which otherwise had two entries and representation of significance by two bits. If the significance test result does not match the pattern, we transmitted the result of the significance test for each entry in the atom. In Refinement pass for each entry in the LSP, except those included in the last sorting pass , output nth bit of the entry [6]. There are two passes in SPIHT one is sorting pass which is initial step and other is refinement pass. In sorting pass sorting is done by comparing two elements at a time, and each comparison results in yes/no. it checks the significance of coefficients present in LIS. If the coefficients are significant then it results in yes and move to LSP. If they are not significant it results in no. In refinement pass it is performed after sorting pass the significant coefficients which we get from sorting pass are send to decoder[8]. 5. Region of Interest (ROI)

Region of interest is the selected portion of the image which contains the information that is required. ROI is a feature introduced to overcome the loss of information in parts of an image which are more important than others [7]. ROI can be defined by a user and they are encoded with better quality than the rest of the image [8]. 6. Medical Images

Medical science grows very fast and hence each hospital needs to store high volume of data about the patients. And medical images are one of the most important data about patients. Medical images are important as they are used by doctors in order to keep record of patients for long term. In order to keep the record of patients for long terms they are compressed using compression techniques so that large amount of data can be store. There are many types of medical images that are used to detect disease of patients. MRI is magnetic resonance image which is used to get information about tissues, organs in human body. Other types are X-ray, CT(computer tomography) , ECG(electrocardiogram) [8]. II. PROPOSED SCHEME

Purposed algorithm modifies the existing SPIHT algorithm with multi wavelet transformation and multi wavelet decomposition will be performed with MFHWT. To perform the operation of compression using improved SPIHT, following algorithm is used:

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME Step1: Read the image as matrix. Step2: Select the region of interest (ROI) that provides the information which is required. Step3: Apply SPHIT algorithm to find the list of significant and insignificant pixels or frequency bands. Step4: To find the LSP we use the multi wavelet decomposition which will perform with the help of MFHWT. Step5: After applying MFHWT we get a transformed image of input image. Step6: for reconstruction process applies the inverse. Step7: Calculate Compression ration and PSNR for reconstructed image. In objective measures of image quality metrics, some statistical indices are calculated to indicate the quality of reconstructed image. The image quality provides some measure of closeness between two digital images by exploiting the differences in the statistical distribution of pixel values. The most commonly used error metrics used for comparing compression are Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR). PSNR computes Peak Signal to Noise Ratio, in decibels, between two images. This ratio is used to provide the quality measurement between the original and a compressed image. Higher the PSNR more will be the quality. MSE computes Mean Square Error between the compressed image and original image. Lower the value of MSE lowers the error. III. EXPERIMENTAL RESULTS

Fig. 2 GUI for image compression

4

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME

Fig. 3 Compression of medical image using SPIHT

Fig. 4 Compression of medical image using ISPIHT

5

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME

Fig. 5 Compression of Teeth using SPIHT

Fig. 6 Compression of Teeth using ISPIHT

6

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME IV. RESULTS AND CALCULATIONS

After the experiments performed on Images in MATLAB, we have realized that above factors determine the quality of reconstructed image. Our technique is better than those of other techniques of compression, because this technique provides better quality, avoid loss of information. The quality of image is measured by the Peak Signal to Noise Ratio (PSNR). Following table provides the PSNR values on the image. Table1. Calculation of PSNR values by applying SPIHT and Improved SPIHT. SR. NO. 1. Techniques PSNR value SPIHT 60.95 Bpp Compression Compressed Ratio Image 3.3787

0.8108

2.

ISPIHT

77.0031

2.5759

10.7328

V.

CONCLUSION

A number of techniques have been proposed on compression; however our proposed technique is better than other techniques as this technique provide more quality and less loss of information. The proposed compression scheme is evaluated on the medical images to compress them with better quality so that there is no loss of information. And can be send to doctors without any loss of information with better quality. Our proposed compression scheme is based on ROI that provide the part of image that contains the information which is required. REFERENCES [1] Kaur Navjot, Singh Preeti, (2012), A new method of image compression using improved SPIHT and MFHWT, IJLRST, Vol.1, Pp-124-126. [2] Liu Bo, Wang Jianjun, (2009), Modified SPIHT based image compression algorithm for hardware implementation, IEEE, Pp-572-576. [3] Bell .E Amy, Martin .B Michael, (2001), New image compression techniques using multi wavelet and multi wavelet packets, IEEE, Vol.10, Pp-500-510. [4] Adams Damien, Patterson Halsey, (2006), The haar wavelet transform: Compression and Reconstruction. [5] U. S. Ragupathy, D. Baskar, A. Tamilarasi, (2008), New method of image compression using multiwavelets and set partitioning algorithm, IEEE. [6] Kalpana .E, Sridhar .V, (2012), ECG data compression using SPIHT algorithm and transmission using Bluetooth technology, IJARECE, Vol.1, Pp-21-29.
7

International Journal of Graphics and Multimedia (IJGM), ISSN 0976 6448(Print), ISSN 0976 6456(Online) Volume 4, Issue 1, January - April 2013, IAEME [7] Amin .H, Dehmeshki .J, Dehkordi .M, Firoozbakht .M, Martini .M, Qanadli .SD, Youannic .A, (2010), Compression of digital medical images based on multiple regions of interest, IEEE, Pp-260-263. [8] Gupta Shipra, Sharma Chirag, (2012), A new method of image compression using multi wavelet technique with MFHWT and ROI in SPIHT , IJITEE, Vol.2, Pp-26-27. [9] John Blesswin, Rema and Jenifer Joselin, A Self Recovery Approach using Halftone Images for Medical Imagery System, International journal of Computer Engineering & Technology (IJCET), Volume 1, Issue 2, 2010, pp. 133 - 146, ISSN Print: 0976 6367, ISSN Online: 0976 6375. [10] Mayuri Y. Thorat and Vinayak K. Bairagi, Hybrid Method to Compress Slices of 3D Medical Images, International journal of Electronics and Communication Engineering & Technology (IJECET), Volume 4, Issue 2, 2013, pp. 250 - 256, ISSN Print: 0976- 6464, ISSN Online: 0976 6472. [11] Rohini N. Shrikhande and Vinayak K. Bairagi, Prediction Based Lossless Medical Image Compression, International journal of Electronics and Communication Engineering & Technology (IJECET), Volume 4, Issue 2, 2013, pp. 191 - 197, ISSN Print: 0976- 6464, ISSN Online: 0976 6472.