Dh33653657

161 views

Published on

International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.

Published in: Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
161
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Dh33653657

  1. 1. Prof. Mr. ArjunNichal, Prof. Mr. PradnyawantKalamkar, / International Journal ofEngineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.653-657653 | P a g eA Novel Approach to Medical & Gray Scale Image EnhancementProf. Mr. ArjunNichal*, Prof. Mr. PradnyawantKalamkar**, Mr.AmitLokhande***, Ms. VrushaliPatil****, Ms.BhagyashriSalunkhe*****Department of Electronics& Telecomm, Adarsh Institute of technology & Research Centre vita,Maharashtra 415311, IndiaABSTRACTImage enhancement is a mean as theimprovement of an image appearance byincreasing dominance of some features or bydecreasing ambiguity between different regions ofthe image. Image enhancement processes consistof a collection of techniques that seek to improvethe visual appearance of an image or to convertthe image to a form better suited for analysis by ahuman or machine. Many images such as medicalimages, remote sensing images, electronmicroscopy images and even real lifephotographic pictures, suffer from poor contrast.Therefore it is necessary to enhance the contrast.The purpose of image enhancement methods is toincrease image visibility and details. Enhancedimage provide clear image to eyes or assistfeature extraction processing in computer visionsystem. Numerous enhancement methods havebeen proposed but the enhancement efficiency,computational requirements, noise amplification,user intervention, and application suitability arethe common factors to be considered whenchoosing from these different methods for specificimage processing applicationKeywords-Image Enhancement, Image Negation,Histogram Equalization, DWT, BPHE.INTRODUCTIONEnhancement is simple and most appealingarea among all the digital image processingtechniques. The main purpose of imageenhancement is to bring out detail that is hidden inan image or to increase contrast in a low contrastimage. Whenever an image is converted from oneform to other such as digitizing the image someform of degradation occurs at output. Imageenhancement is among the simplest and mostappealing areas of digital image processing.Basically, the idea behind enhancement techniquesis to bring out detail that is obscured, or simply tohighlight certain features of interest in an image.Enhanced images provide better contrast of thedetails that images contain. Image enhancement isapplied in every field where images are ought to beunderstood and analyzed. For example, MedicalImage Analysis, Analysis of images from satellites,etc. Image enhancement is among the simplest andmost appealing areas of digital imageprocessing.Basically, the idea behind enhancement techniquesis to bring out detail that is obscured, or simply tohighlight certain features of interest in an image.Why we are moving towards imageenhancement?Enhanced images provide better contrast ofthe details that images contain. Image enhancementis applied in every field where images are ought tobe understood and analyzed. For example, MedicalImage Analysis, Analysis of images from satellites,etc.There are three types of image enhancementwhich are as follows:a. Spatial domainFig 1: Spatial domainSpatial domain approaches direct information ofpixel in an image. The image processing functionmay be expressed as :G(x,y)=T{f(x,y)}where f(x,y) is the inputimage & g(x,y) is the proposed image.b. Frequency domainIn frequency domain the Fourier transform of animage is modified. Fourier series: Any function thatperiodically repeats itself can be expressed as thesum of sine‟s /cosines of different frequencies, eachmultiplied with a different coefficient. FourierTransform: Functions that are not periodic, whosearea under the curve is finite, can be expressed asthe integral of sine‟s and/cosines multiplied by aweighting function.c. Transform domainTransforming image intensity data into specificdomain includes altering high-frequency content ofimage. Using discrete cosine, Fourier, and wavelettransformsI. MAIN METHODOLOGYa. Image Negation MethodNow contrast and poor quality are mainproblems in the production of medical images.
  2. 2. Prof. Mr. ArjunNichal, Prof. Mr. PradnyawantKalamkar, / International Journal ofEngineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.653-657654 | P a g eMedical image enhancement technologies haveattracted much attention since advanced medicalequipment‟s were put into use in the medical field.Negation method is generally used to enhance themedical image. Negation of image is nothing butreversing the intensity levels of an image to producethe equivalent of photographic negative. This typeof processing is particularly suited for enhancingwhite or gray details embedded in dark region of animage especially when the black areas are dominantin size. The negative point transformation functionalso known as contrast reverse. The negativetransformation shown in fig. is obtained byfollowing expression, s=L-1-rWhere„s‟ is Output image after transformationL-1 is Maximum Pixel valuer -is Input Image.Fig 2: Negative Image graphb. Histogram Equalization Method (HE)Histogram equalization is a technique bywhich the gray-level distribution of an image ischanged in such a way as to obtain a uniform (flat)resulting histogram, in which the percentage ofpixels of every gray level is the same. To performhistogram equalization, it is necessary to use anauxiliary function, called the transformationfunction, T (r). Such transformation function mustsatisfy two criteria1. T (r) must be a monotonically increasing functionin the interval 0 ≤ r ≤ L − 1.2. 0 ≤ T (r) ≤ L − 1 for 0 ≤ r ≤ L − 1.The most usual transformation function is thecumulative distribution function (cdf) of the originalprobability mass function, given by Histogramequalization is used for increasing contrast of animage. This can be achieved by using histogramstretching operation [3].Fig 3: Histogram of Original ImageFig 4: Histogram of Enhanced ImageFig.3 shows histogram of original image, when weapply the Histogram Equalization method we get theenhanced image. Fig.4 shows stretched histogram ofenhanced image.C. DWT (Discrete Wavelet Transform) basedMethodAerial images captured from aircrafts,spacecraft‟s, or satellites usually suffer from lack ofclarity, since the atmosphere enclosing Earth haseffects upon the images such as turbidity caused byhaze, fog, clouds or heavy rain. The visibility ofsuch aerial images may decrease drastically andsometimes the conditions at which the images aretaken may only lead to near zero visibility even forthe human eyes. Even though human observers maynot see much than smoke, there may exist usefulinformation in those images taken under such poorconditions[1].Recently we use a wavelet-based dynamicrange compression algorithm to improve the visualquality of digital images captured in the highdynamic range scenes with no-uniform lightingconditions. The fast image enhancement algorithmwhich provides dynamic range compressionpreserving the local contrast and tonal rendition is avery good candidate in aerial imagery applicationssuch as image interpretation for defense and In thispaper the latest version of the proposed algorithmwhich is able to enhance aerial images so that theenhanced images are better than direct humanobservation, is presented. The results obtained byapplying the algorithm to numerous aerial imagesshow strong robustness and high image quality.The proposed enhancement algorithm consists ofthree stages. The first and the third stage are applied
  3. 3. Prof. Mr. ArjunNichal, Prof. Mr. PradnyawantKalamkar, / International Journal ofEngineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.653-657655 | P a g ein the spatial domain and the second one in thediscrete wavelet domain.Fig 5: Flow chart of DWTD. BPHE Method (Brightness Preserving BiHistogram Equalization):The Brightness preserving bi histogramequalization firstly decomposes an input image intotwo sub images based on the mean of the inputimage. One of the sub image is set of samples lessthan or equal to the mean whereas the other one isthe set of samples greater than the mean. Then theBBHE equalizes the sub images independentlybased on their respective histograms with theconstraint that the samples in the formal set aremapped into the range from the minimum gray levelto the input mean and the samples in the latter setare mapped into the range from the mean t themaximum gray level. Means one of the sub imagesis equalized over the range up to the mean and theother sub image is equalized over the range. Fromthe mean based on the respective histograms .Thus,the resulting equalized sub images are bounded byeach other around the input mean, which has aneffect of preserving mean brightness[2].Fig 6. Flow chart of BPHEII. QUALITY PARAMETERSDepending on the parameter value we can determinein what extent an image is enhanced.1. The MSE between two images f and g isdenoted by,𝑀𝑆𝐸 =1𝑀𝑁(𝑓 𝑗, 𝑘 − 𝑔(𝑗, 𝑘))2𝑗 ,𝑘Where the sum over j; k denotes the sum over allpixels in the images, and m is the number of rows, nis the number of column of each image.2. The PSNR between two (8 bpp) images is, indecibels,𝑃𝑆𝑁𝑅 = 10log⁡2552𝑀𝑆𝐸PSNR tends to be cited more often, since it is alogarithmic measure, and our brains seem torespond logarithmically to intensity. IncreasingPSNR represents increasing fidelity of compression.Generally, when the PSNR is 40 dB or larger, thenthe two images are virtually indistinguishable byhuman observers.3. Structural Content (SC)Structural Content is defined as,𝑆𝐶 =𝑀, 𝑁[𝐼1 𝑚, 𝑛 .∗ 𝐼1(𝑚, 𝑛)]𝑀, 𝑁[𝐼2 𝑚, 𝑛 .∗ 𝐼2(𝑚, 𝑛)]The large value of Structural Content (SC) meansthat image is of poor quality.4. Average Difference (AD)Average Difference (AD) is defined as:𝐴𝐷 =𝑀, 𝑁[𝐼1 𝑚, 𝑛 − 𝐼1 𝑚, 𝑛 ]𝑀 ∗ 𝑁The large value of AD means that the pixel values inthe reconstructed image are more deviated fromactual pixelvalue. Larger value of AD indicatesimage is of poor quality.5. Absolute means brightness error (AMBE):It is the Difference between original and enhancedimage and is given asAMBE=E(x)-E(y)Where E(x)= average intensity of input imageE(y)=average intensity of enhanced image6. Contrast:Contrast defines the difference between lowest andhighest intensity level. Higher the value of contrastmeans more difference between lowest and highestintensity level.Wavelet based Dynamic range compression& Contrast EnhancementHistogram AdjustmentColor RestorationSTOPSTARTInitialize the imageFind mean or medianMake two parts: 1)0-mean2)mean+1-maxFind Histogram of each partCombine the HistogramSTOP
  4. 4. Prof. Mr. ArjunNichal, Prof. Mr. PradnyawantKalamkar, / International Journal ofEngineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.653-657656 | P a g eIII. RESULT ANALYSISFollowing figures shows original images and theirenhanced images.Fig 7. Original Lena ImageFig 8. Image Negation of Lena ImageFig 9. Enhanced Lena Image using HEFig 10. Histogram of Original LenaFig 11. Histogram of Negative Lena ImageFig 12. Histogram of Enhanced Image0 50 100 150 200 250 3000500100015002000250030003500histogram0 50 100 150 200 250 300050010001500200025003000histogram0 50 100 150 200 250 3000100020003000400050006000histogram
  5. 5. Prof. Mr. ArjunNichal, Prof. Mr. PradnyawantKalamkar, / International Journal ofEngineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.653-657657 | P a g eFig 13. Blurred Color Port ImageThe performances of these techniques are evaluatedin terms of PSNR, AMBE, CONTRAST,MSE, SCand AD.Table 1: Analysis of Different methods withvarious parametersIV. CONCLUSIONSIn this paper different image enhancementtechniques are used to enhance the images fromdifferent area. Such as Negative image enhancementis used to enhance the images in medical field. HE isused to enhance the images which are captured fromdigital camera. DWT is used to enhance the Arielimages, e.g. Images captured through satellite orspacecraft .It is used to enhance the color images.BPHE technique is advanced version of HE. Itincreases the contrast of an image better than HE.REFERENCES[1] AnamikaBhardwaj& Manish K.Sing “ANovel approach of medical imageenhancement based on Wavelet transform”Vol. 2, Issue 3, May-Jun 2012, pp.2356-2360[2] Rajesh Garg, Bhawna Mittal&SheetalGarg, “Histogram EqualizationTechniques For Image Enhancement”IJECT Vol. 2, Issue 1[3] S. Lau, “Global image enhancement usinglocalinformation,” Electronics Letters, vol.30, pp. 122–123,Jan. 1994.[4] J. Zimmerman, S. Pizer, E. Staab, E. Perry,W. McCartney,B. Brenton, “Evaluation ofthe effectiveness of adaptivehistogramequalization”.Fig 14. Enhanced Port Image using DWT method[5]Digital image processing by Madhuri A. Joshi.page no. 69-96[6] MATLAB and applications in engg. By Rajkumarbansal, Ashok kumarGoel, Manoj Kumar .Prof. ArjunNichal working asa Assistant professor in AITRCvita. Received M.Tech inelectronics from WalchandCollege of engineering, Sangli,His area of interest is DigitalImage Processing, DigitalSignal Processing and Embedded system.Prof. PradnyawantKalamkarworking as a Assistant professorin AITRC vita. ReceivedM.Tech in electronics fromWalchand College ofengineering, SangliHis area ofinterest is Digital ImageProcessing, and Wireless communication.Mr. AmitLokhandepursuinghis B.E degree in Electronicsand telecommunication fromAITRC, vita. His area of interestis Digital Image Processing.Ms. VrushaliPatil pursuing herB.E degree in Electronics andtele. from AITRC, vita. Her areaof interest is Digital ImageProcessing,Ms. BhagyashriSalunkhepursuing her B.E degree inElectronics and tele. fromAITRC, vita. Her area ofinterest is Digital ImageProcessing, and Embeddedsystem.

×