ISSN : 2230-7109(Online) ISSN : 2230-9543(Print) IJECT .
IJECT Vol. 2, Issue 1, March 2011ISSN : 2230-7109(Online) ISSN : 2230-9543(Print)Histogram Equalization Techniques ForImage Enhancement1Rajesh Garg, 2Bhawna Mittal, 3Sheetal GargH.I.T., Sonepat, Haryana, IndiaS.M.Hindu Sr.Sec.School, Sonepat, Haryana, India1,23AbstractVarious enhancement schemes are used for enhancing animage which includes gray scale manipulation, filtering andHistogram Equalization (HE). Histogram equalization is oneof the well known imaget enhancement technique. It becamea popular technique for contrast enhancement because thismethod is simple and effective. In the latter case, preservingthe input brightness of the image is required to avoid thegeneration of non-existing artifacts in the output image.Although these methods preserve the input brightness on theoutput image with a significant contrast enhancement, theymay produce images with do not look as natural as the inputones. The basic idea of HE method is to re-map the gray levelsof an image. HE tends to introduce some annoying artifactsand unnatural enhancement. To overcome these drawbacksdifferent brightness preserving techniques are used whichare covered in the literature survey. Comparative analysis ofdifferent enhancement techniques will be carried out. Thiscomparison will be done on the basis of subjective and objectiveparameters. Subjective parameters are visual quality andcomputation time and objective parameters are Peak signalto-noise ratio (PSNR), Mean squared error (MSE), NormalizedAbsolute Error (NAE), Normalized Correlation, Error Color andComposite Peak Signal to Noise Ratio (CPSNR).KeywordsContrast enhancement, Histogram equalization, PSNR, MSE,NAE, CPSNR, Visual quality.I. IntroductionOut of the five senses – sight, hearing, touch, smell and taste– which humans use to perceive their environment, sight isthe most powerful. Receiving and analyzing images forms alarge part of the routine cerebral activity of human beingsthroughout their waking lives. In fact, more than 99% of theactivity of the human brain is involved in processing imagesfrom the visual cortex. A visual image is rich in information.Confucius said, “A picture is worth a thousand words.” [1] ImageEnhancement is simple and most appealing area among allthe digital image processing techniques. The main purpose ofimage enhancement is to bring out detail that is hidden in animage or to increase contrast in a low contrast image. Wheneveran image is converted from one form to other such as digitizingthe image some form of degradation occurs at output.A. Image EnhancementImage enhancement is among the simplest and most appealingareas of digital image processing. Basically, the idea behindenhancement techniques is to bring out detail that is obscured,or simply to highlight certain features of interest in an image.A familiar example of enhancement is shown in Fig.1 in whichwhen we increase the contrast of an image and filter it to removethe noise "it looks better." It is important to keep in mind thatenhancement is a very subjective area of image processing.Improvement in quality of these degraded images can beachieved by using application of enhancement techniques.w w w. i j e c t. o r gFig.1: Image enhancementB. Adaptive Histogram Equalization methodThis is an extension to traditional Histogram Equalizationtechnique. It enhances the contrast of images by transformingthe values in the intensity image I. Unlike HISTEQ, it operateson small data regions (tiles), rather than the entire image. Eachtile's contrast is enhanced, so that the histogram of the outputregion approximately matches the specified histogram. Theneighboring tiles are then combined using bilinear interpolationin order to eliminate artificially induced boundaries.The contrast, especially in homogeneous areas, can be limitedin order to avoid amplifying the noise which might be presentin the image.C. Dualistic sub-image histogram equalization methodThis is a novel histogram equalization technique in which theoriginal image is decomposed into two equal area sub-imagesbased on its gray level probability density function.Then the two sub-images are equalized respectively. At last, weget the result after the processed sub-images are composedinto one image. In fact, the algorithm can not only enhancethe image visual information effectively, but also constrain theoriginal image's average luminance from great shift. This makesit possible to be utilized in video system directly.C. Dynamic histogram equalization for image contrastenhancementIt employs a partitioning operation over the input histogramto chop it into some sub histograms so that they have nodominating component in them. Then each sub-histogram goesthrough HE and is allowed to occupy a specified gray level rangein the enhanced output image. Thus, a better overall contrastenhancement is gained by DHE with controlled dynamic rangeof gray levels and eliminating the possibility of the low histogramcomponents being compressed that may cause some part ofthe image to have washed out appearance.II. BackgroundOne of the first applications of digital images was in thenewspaper industry, when pictures were first sent by submarinecable between London and New York. Introduction of theBartlane cable picture transmission system in the early 1920sreduced the time required to transport a picture across theAtlantic from more than a week to less than three hours.Specialized printing equipment coded pictures for cabletransmission and then reconstructed them at the receiving end.International Journal of Electronics & Communication Technology107
IJECT Vol. 2, Issue 1, March 2011Some of the initial problems [2] in improving the visual qualityof these early digital pictures were related to the selection ofprinting procedures and the distribution of intensity levels.Although the methods just cited involve digital images, theyare not considered digital image processing results in thecontext of our definition because computers were not involvedin their creation. Thus, the history of digital image processingis intimately tied to the development of the digital computer. Infact, digital images require so much storage and computationalpower that progress in the field of digital image processing hasbeen dependent on the development of digital computers andof supporting technologies that include data storage, display,and transmission.III. ImplementationCompare all these techniques on the basis of performanceparameters in objective and subjective manner. These arethe merits on the bases of that I will compare above definedtechniques.A. Contrast Limited Adaptive Histogram Equalizationmethod:Algorithm Steps:1. Obtain all the inputs: Image, Number of regions in row andcolumn directions, Number of bins for the histograms usedin building image transform function (dynamic range), Cliplimit for contrast limiting (normalized from 0 to 1)2. Pre-process the inputs: Determine real clip limit fromthe normalized value if necessary, pad the image beforesplitting it into regions3. Process each contextual region (tile) thus producing graylevel mappings: Extract a single image region, make ahistogram for this region using the specified number ofbins, clip the histogram using clip limit, create a mapping(transformation function) for this region4. Interpolate gray level mappings in order to assemble finalCLAHE image: Extract cluster of four neighbouring mappingfunctions, process image region partly overlapping eachof the mapping tiles, extract a single pixel, apply fourmappings to that pixel, and interpolate between the resultsto obtain the output pixel; repeat over the entire image.ISSN : 2230-7109(Online) ISSN : 2230-9543(Print)B. Equal area Dualistic sub-image histogram equalizationmethod:Algorithm Steps:Suppose image X is segmented by a section with gray levelof X Xe and the two sub-images are Xl and Xu, so we haveX XlUִ XU. HereX L (1)It is obvious that sub image XL is composed by gray level of{X 0,X1, Xe-1}, while sub image X U is composed of{Xe,Xe 1, .Xl-1}. The aggregation of the original images’ graylevel distribution probability is decomposed into {p0,p1, pe-1}and {pe,pe 1, .pl-1} correspondingly. The correspondingcumulative distribution function will beCL( X k) , k 0 ,1 , .e-1.(2)CU (Xk) , k e,e 1, . L-1Based on the cumulative distribution function, the transformfunctions of the two sub images’ histogram are equalizedbelow.F L( X k) X 0 ( X e-1- X 0) c ( X k) , k 0 , 1 , . e - 1 (3)FU(Xk) Xe (Xl-1-Xe)c(Xk), k e,e 1, .L-1At last result of dualistic sub image histogram is obtained afterthe two equalized sub images are composed into one image.Suppose Y denotes the processed image thenY {Y(i,j)} f L (X L ) Uִf U (X U ) .(4)Fig.3: Flow chart for DSIHEFig.2: Flow chart for CLAHE108International Journal of Electronics & Communication Technologyw w w. i j e c t. o r g
IJECT Vol. 2, Issue 1, March 2011ISSN : 2230-7109(Online) ISSN : 2230-9543(Print)C. Dynamic histogram equalization for image contrastenhancement:Algorithm Steps:1. Histogram Partition : DHE partitions the histogram based onlocal minima. At first, it applies a one-dimensional smoothingfilter of size 1 x 3 on the histogram to get rid of insignificantminima. Then it makes partitions (sub-histograms) taking theportion of histogram that falls between two local minima (thefirst and the last non-zero histogram components are consideredas minima). Mathematically, if m0, m1, , mn are (n 1) graylevels (GL) that correspond to (n 1) local minima in the imagehistogram, then the first sub-histogram will take the histogramcomponents of the GL range [m0, m1], the second one will take[m1 1, m2] and so on.These histogram partitioning helps toprevent some parts of the histogram from being dominatedby others.2. Gray Scale Allocation: For each sub-histogram, DHE allocatesa particular range of GLs over which it may span in outputimage histogram. This is decided mainly based on the ratio ofthe span of gray levels that the sub-histograms occupy in theinput image histogram.Here the straightforward approach isSpani mi-mi-13. Contrast:Contrast defines the difference between lowest and highestintensity level. Higher the value of contrast means moredifference between lowest and highest intensity level.4. Visual QualityBy looking at the enhanced image, one can easily determinethe difference between the input image and the enhancedimage and hence, performance of the enhancement techniqueis evaluated.rangei where, spani dynamic GL range used by sub-histogram i ininput image.mi ith local minima in the input image histogram.rangei dynamic gray level range for sub-histogram i in outputimage.The order of gray levels allocated for the sub-histograms inoutput image histogram are maintained in the same order asthey are in the input image, i.e., if sub-histogram i is allocatedthe gray levels from [istart, iend], then istart (i-1)end 1 and iend istart rangei. For the first sub-histogram, j, jstart r0.3. Histogram Equalization : Conventional HE is applied to eachsub-histogram, but its span in the output image histogramis allowed to confine within the allocated GL range that isdesignated to it. Therefore, any portion of the input imagehistogram is not allowed to dominate in HE.D. Metrics for Gray Scale Images:Fig. 4: Flow chart for DHE1. Peak-signal-to-noise-ratio (PSNR):PSNR is the evaluation standard of the reconstructed imagequality, and is important measurement feature. PSNR ismeasured in decibels (dB) and is given by:IV. Tool to be used:In this thesis for implementation of techniques MATLAB 7.0.2version is used. In that image processing toolbox is used.MATLAB is a high-performance language for technicalcomputing. 2 PSNR 10log 255 MSE where the value 255 is maximum possible value that canbe attained by the image signal. Mean square error (MSE) isdefined as Where M*N is the size of the original image. Higherthe PSNR value is, better the reconstructed image is.2. Absolute mean brightness error (AMBE):It is the Difference between original and enhanced image andis given asAMBE Where E(x) average intensity of input image E(y) averageintensity of enhanced imagew w w. i j e c t. o r gV. Experimental ResultsTo verify the efficacy of the proposed method, obtained afterfollowing the Different enhancement Algorithms for gray scaleimages. After the comparison tables, a graphical representationhas also been done for a quick analysis of results. All thetechniques have been tested for all the assumed standard testimages.In this paper three techniques are used for Gray ScaleImage enhancement which are CLAHE, DSIHE and DHE.A. Results of test image “Rice”Fig.5 shows the visual quality of real image “Rice” and theenhanced image using three different image enhancementtechniques. The performances of these techniques areevaluated in terms of PSNR, AMBE and Contrast.International Journal of Electronics & Communication Technology109
IJECT Vol. 2, Issue 1, March 2011ISSN : 2230-7109(Online) ISSN : 2230-9543(Print)(a) Histogram of Imagea) Original Imageb) CLAHE Image(b) CLAHE Histogramc) DSIHE Image(c ) Histogram of DHE(d) DSIHE Histogramd) DHE ImageFig. 5: Enhanced Result of real image as shown in image a,b, c, d.B. Histograms of test image “Rice” for differentenhancement algorithmsFig.6 shows respective Histograms of test image “rice” usingDifferent image enhancement techniques.110International Journal of Electronics & Communication TechnologyFig. 6: Equalized Histograms for Image “Rice” as shown inimage a, b, c, d as original, CLAHE, DHE, DSIHE Respectively.Table 1: Comparison of Various Parameters for “Rice” Image:Parameter AMBEContrast PSNRTechniqueCLAHE13.852123.5878 0.0366DSIHE4.908133.8767 0.0327DHE13.0886 12.1438 0.1107Anyone can make comparison of parameter AMBE (Absolutemean brightness error) for different image enhancementtechniques. The value of AMBE should be as small as possiblew w w. i j e c t. o r g
ISSN : 2230-7109(Online) ISSN : 2230-9543(Print)which indicates that difference between original and enhancedimage should be minimum. Therefore in terms of AMBE, DSIHEtechnique gives best results as AMBE is taken in negative.Now considering PSNR, CLAHE gives better output as it is clearedfrom the formula that PSNR should be as high as possible sothat noise content should be lower than signal content.VI. Conclusion and Future ScopeIn this Paper, a frame work for image enhancement basedon prior knowledge on the Histogram Equalization has beenpresented. Many image enhancement schemes like Contrastlimited Adaptive Histogram Equalization (CLAHE), Equal areadualistic sub-image histogram equalization (DSIHE), DynamicHistogram equalization (DHE) Algorithm has been implementedand compared. The Performance of all these Methods has beenanalyzed and a number of Practical experiments of real timeimages have been presented. From the experimental results, itis found that all the three techniques yields Different aspects fordifferent parameters. In future, for the enhancement purposemore images can be taken from the different application fieldsso that it becomes clearer that for which application whichparticular technique is better both for Gray Scale Images andcolour Images. Particularly, for colour images there are notmany performances measurement parameter considered.So, new parameters can be considered for the evaluation ofenhancement techniques. New colour models can also bechosen for better comparison purpose. Optimization of variousenhancement techniques can be done to reduce computationalcomplexity as much as possible.References[1] S. Lau, “Global image enhancement using localinformation,” Electronics Letters, vol. 30, pp. 122–123,Jan. 1994.[2] J. Zimmerman, S. Pizer, E. Staab, E. Perry, W. McCartney,B. Brenton, “Evaluation of the effectiveness of adaptivehistogram equalization for contrast enhancement,” IEEETransactions on Medical Imaging, pp. 304-312, 1988.[3] Yu Wan, Qian Chen, Bao-Min Zhang, “Image enhancementbased on equal area dualistic sub-image histogramequalization method,” IEEE Transactions ConsumerElectron., vol. 45, no. 1, pp. 68-75, 1999.[4] Yeong-Taeg Kim, “Contrast enhancement using brightnesspreserving bi-histogram equalization,” IEEE Trans.Consumer Electronics, vol. 43, no. 1, pp. 1-8, 1997.[5] M. Abdullah-Al-Wadud, Md. Hasanul Kabir, M. Ali AkberDewan, Oksam Chae, “A dynamic histogram equalizationfor image contrast enhancement”, IEEE Transactions.Consumer Electron., vol. 53, no. 2, pp. 593- 600, May2007.[6] K. Wongsritong, K. Kittayaruasiriwat, F. Cheevasuvit,K. Dejhan, A. Somboonkaew, “Contrast enhancementusing multipeak histogram equalization with brightnesspreserving”, Circuit and System, 1998, IEEE APCCAS1998. The 1998 IEEE Asia-Pacific Conference on 24-27Nov. 1998, pp. 455-458, 1998.[7] Y. Wang, Q. Chen, B. Zhang, Soong-Der Chen, and Abd.Rahman Ramli, “Minimum mean brightness error bihistogram equalization in contrast enhancement”, IEEETransactions Consumer Electron. vol. 49, no. 4, pp. 13101319, Nov. 2003.[8] WANG Zhiming, TAO Jianhua, “A Fast Implementation ofAdaptive Histogram Equalization”, IEEE 2006, ICSP 2006w w w. i j e c t. o r gIJECT Vol. 2, Issue 1, March 2011Proceedings.[9] Md. Foisal Hossain, Mohammad Reza Alsharif, “ImageEnhancement Based on Logarithmic Transform Coefficientand Adaptive Histogram Equalization”, 2007 InternationalConference on Convergence Information Technology, IEEE2007.[10] J. Alex Stark “Adaptive Image Contrast EnhancementUsing Generalizations of Histogram Equalization”, IEEETransactions on Image Processing, Vol. 9, No. 5, May2000.[11] Wang Yuanji. Li Jianhua, Lu E, Fu Yao, Jiang Qinzhong,“Image Quality Evaluation Based On Image WeightedSeparating Block Peak Signal To Noise Ratio”, IEEE Int.Conf. Neural Networks & Signal Processing, Nanjing,China, December 14-17, 2003.[12] Rafael C. Gonzalez, Richard E. Woods, “Digital ImageProcessing”, 2nd edition, Prentice Hall, 2002.[13] Stephen M. Pizer, R. Eugene Johnston, James P. Ericksen,Bonnie C. Yankaskas, Keith E. Muller, “Contrast-LimitedAdaptive Histogram Equalization Speed and Effectiveness”,”, IEEE Int. Conf. Neural Networks & Signal Processing,Nanjing, China, December 14-17, 2003.[14] Rafael C. Gonzalez, Richard E. Woods, “Digital ImageProcessing”, 2nd edition, Prentice Hall, 2002.[15] A. K. Jain, “Fundamentals of Digital Image Processing”.Englewood Cliffs, NJ: Prentice-Hall, 1991.[16] A. Zagzebski, “Essentials of Ultrasound Physics”. St. Louis,Missouri: Mosby, 1996.Rajesh Garg received his B.E. degree inElectronics & Comm. from Hindu College ofengineering, Sonipat, Haryana, in 2006 andpursuing the M-Tech. (part-time) degree inElectronics & Comm. From M.M. University,Mullna (Ambala). Presently, he is engagedin teaching, as a lecturer in Electronics &Comm. Department in Hindu institute ofTechnology, Sonepat since 2006.Bhawna Mittal received her B.E. degreein Electronics & Comm. from NorthMaharashtra University in 1998 andM-Tech. (part-time) degree in Electronics &Comm. from Rajasthan University, Udaipurin 2007. She was teaching as lecturerin S.J.P.P.,Damla in 1999 to 2000.From2000 onwards she worked as a Lecturerin Electronics & Comm. Department atHindu institute of Technology, Sonepat then promoted as Sr.Lecturer in Electronics & Comm. Department at Hindu instituteof Technology, Sonepat in 2007.Presently,she is engaged in allthe academic activities of the institute.Sheetal Garg received her B.Sc. degreein Computer Science from G.V.M. GirlsCollege, Sonipat, Haryana, in 2002 andM.C.A degree from Kurukshetra University,Kurukshetra in 2005. Presently, sheis engaged in teaching, as a lecturerin Computer Science Department inS.M.Hindu Sr.Sec.School, Sonepat since2005.International Journal of Electronics & Communication Technology111
1Rajesh Garg, 2Bhawna Mittal, 3Sheetal Garg 1,2H.I.T., Sonepat, Haryana, India 3S.M.Hindu Sr.Sec.School, Sonepat, Haryana, India In t e r n a t I o n a l Jo u r n a l o f el e c t r o n I c s & co m m u n I c a t I o n te c h n o l o g y 107 ISSN : 2230-7109(Online) ISSN : 2230-9543(Print) IJECT Vo l. 2
IJECT Vo l. 5, Issu E spl-1, Jan - Mar C h 2014 ISSN : 2230-7109 (Online) ISSN : 2230-9543 (Print) 70 InternatIonal Journal of electronIcs & communIcatIon technology www.iject.org Adiabatic Technique for Power Efficient Logic Circuit Design 1Anu Priya, 2Amrita Rai 1,2Dept. of Electronics and Communication, RIET, Haryana, India Abstract The Power dissipation in conventional CMOS circuits can be
IJECT Vo l. 8, Is s u E 3, Ju l y - sE p T 2017 www.iject.org InternatIonal Journal of electronIcs & communIcatIon technology 31 Issn : 2230-7109 (online) Issn : 2230-9543 (Print) Design of 4 Bit FLASH Analog to Digital Converter Using TM Comparator Circuit and Gray to Base2 Encoder us
Security One security-cable slot (wedge-shaped) Card slot microSD card slot SIM card uSIM card slot Table 8. Internal ports and connectors Description Values Internal: One M.2 Key-M (2280 or 2230) for solid-state drive One M.2 3042 Key-B for WWAN One M.2 2230 Key-E for WLAN One M.2 2230 slot for solid-state drive 128 GB/256 GB/512 GB
Product Data Sheet April 2018 00813-0100-2230, Rev BB Rosemount 2230 Graphical Field Display for tank gauging systems Monitor real time tank gauging data, such as level, temperature, pressure, and total observed volume
Your Alienware m17 R4 supports one of the following storage configurations: One M.2 2230/2280 solid-state drive Two M.2 2230 solid-state drives Two M.2 2280 solid-state drives One M.2 2230 solid-state drive and two M.2 2280 solid-state drives The primary drive of your Alienware m17 R4 varies with the storage configuration.
Required if volunteer will be driving a government vehicle per FSM 7109.19_61.1. Defensive driving training required within last 3 years. International driver licenses are not valid to drive a Forest Service vehicle on most units, per FSM 7109.
Slots 1 SD 3.0 card slot 1 M.2 2230 slot for Wi-Fi and Bluetooth card 1 M.2 2230/2280 slots for SSD/Intel Optane 1 Kensington security-cable slot Number of Bays One internal 2.5 inch Primary Hard Drive1,5 2.5-inch, 500 GB, 5400 RPM, SATA HDD 2.5-inch, 1
Academic literary criticism prior to the rise of “New Criticism” in the United States tended to practice traditional literary history: tracking influence, establishing the canon of major writers in the literary periods, and clarifying historical context and allusions within the text. Literary biography was and still is an important interpretive method in and out of the academy; versions of .
