an improved cellular automata (ca) based image denoising ...€¦ · an improved cellular automata...

9
An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1* , Malathi P 2 , Nagarani S 3 , Oswalt Manoj S 4 1 Department of Information Technology, Sri Ramakrishna Institute of Technology, Coimbatore, India 2 Sri Ranganathar Institute of Engineering and Technology, Coimbatore, India 3 Department of Science and Humanities, Sri Ramakrishna Institute of Technology, Coimbatore, India 4 Department of Computer Science and Engineering, Sri Ramakrishna Institute of Technology, Coimbatore, India Abstract Biometric recognition plays a more important role in the industries and various plays to allow authenticated entry. The database would be created using the biometrics gathered from the users and authentication will be done by matching user biometric with the database. However this matching might increase the false positive rate in case of presence of salt and pepper noise in the images. This effort antagonizes a completely unique approach for screaming pel recognition and refurbishment of grey scale image victimization mirrored cellular automata (CA). The planned method eliminates salt and pepper noise from a tainted image. The planned technique permits extension lead of window size vigorously throughout great noise concentrations. The planned technique uses reflected CA supported Moore neighborhood (8-neighborhood cell). This image denoising method is examined with numerous prevailing image denoising methods like, Median filter, Switch Median (SM) filter, Directional- Weighted-Median (DWM) filter, changed DWM (MDWM), Fuzzy Cellular Automata (FCA), changed call primarily based asymmetrical cut Median Filter (MDBUTMF), and Contra mean value (CHM) filter. 3 sample pictures (Finger print, Iris, and Mandrill) of 2 completely diverse determinations (512 × 512) and (256 × 256) are engaged for the recital examination. The reflected CA is evaluated in contradiction of Peak signal-to-noise (PSNR), Mean square Error (MSE), and Structural SIMilarity (SSIM). It’s ascertained that reflected CA performs higher than the opposite prevailing practices by means of PSNR. Keywords: Cellular automata (CA), Mirrored CA, Mean squared error (MSE), Peak signal-to-noise ratio (PSNR), and Structural SI milarity (SSIM). Accepted on February 01, 2017 Introduction The furthermost prevalent approaches of protecting data and resources protected are to practice password and User ID/PIN security. These systems necessitate the user to validate them by inward bound a "secret" password that they had beforehand fashioned or were allocated. These schemes are disposed to equitation from either a brute force effort to blow the password or from passwords which were not sole or even which were dispatched adjacent the computer itself. A Biometric Identification scheme is one in which the user's "body" turn out to be the password/PIN. Biometric features about the distinct are what make that person sole and consequently can be cast- off to validate a user's admittance to numerous schemes. Biometrics bids sole assistances for recognizing human beings. Tokens, Photo ID cards can be misplaced, pinched, replicated or left at home, Passwords can be elapsed communal or perceived but Biometrics grips the potential of firm, easy-to- use, perfect, consistent, and less costly authentication for an assortment of applications. The biometric identification methods substitute the token recognition scheme where a person is proved by means of his license or pass codes and favored over outdated passwords and PIN-based approaches. Certain recognized Biometric identification methods are stated underneath, “Fingerprint recognition, Voice/sound recognition, Palm vein recognition, Iris recognition, Retina scan recognition and so on”. Denoising is a crucial issue in image process. Impulse noises area unit evoked as a result of bad pixels in camera sensors or broadcast during a clattering channel. Two common forms of impulse noise area unit the random valued noise and the salt- and-pepper noise. For the photographs tainted by salt-and- pepper noise, clattering pixels take solely the utmost and the least standards within the dynamic vary. These noises will ISSN 0970-938X www.biomedres.info Biomedical Research 2018; Special Issue: S31-S39 S31 Special Section: Computational Life Sciences and Smarter Technological Advancement Biomed Res 2018 Special Issue

Upload: others

Post on 30-Apr-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

An improved cellular automata (ca) based image denoising method forbiometric applications.

Suresh A1*, Malathi P2, Nagarani S3, Oswalt Manoj S4

1Department of Information Technology, Sri Ramakrishna Institute of Technology, Coimbatore, India2Sri Ranganathar Institute of Engineering and Technology, Coimbatore, India3Department of Science and Humanities, Sri Ramakrishna Institute of Technology, Coimbatore, India4Department of Computer Science and Engineering, Sri Ramakrishna Institute of Technology, Coimbatore, India

Abstract

Biometric recognition plays a more important role in the industries and various plays to allowauthenticated entry. The database would be created using the biometrics gathered from the users andauthentication will be done by matching user biometric with the database. However this matching mightincrease the false positive rate in case of presence of salt and pepper noise in the images. This effortantagonizes a completely unique approach for screaming pel recognition and refurbishment of greyscale image victimization mirrored cellular automata (CA). The planned method eliminates salt andpepper noise from a tainted image. The planned technique permits extension lead of window sizevigorously throughout great noise concentrations. The planned technique uses reflected CA supportedMoore neighborhood (8-neighborhood cell). This image denoising method is examined with numerousprevailing image denoising methods like, Median filter, Switch Median (SM) filter, Directional-Weighted-Median (DWM) filter, changed DWM (MDWM), Fuzzy Cellular Automata (FCA), changedcall primarily based asymmetrical cut Median Filter (MDBUTMF), and Contra mean value (CHM)filter. 3 sample pictures (Finger print, Iris, and Mandrill) of 2 completely diverse determinations (512 ×512) and (256 × 256) are engaged for the recital examination. The reflected CA is evaluated incontradiction of Peak signal-to-noise (PSNR), Mean square Error (MSE), and Structural SIMilarity(SSIM). It’s ascertained that reflected CA performs higher than the opposite prevailing practices bymeans of PSNR.

Keywords: Cellular automata (CA), Mirrored CA, Mean squared error (MSE), Peak signal-to-noise ratio (PSNR), andStructural SI milarity (SSIM).

Accepted on February 01, 2017

IntroductionThe furthermost prevalent approaches of protecting data andresources protected are to practice password and User ID/PINsecurity. These systems necessitate the user to validate them byinward bound a "secret" password that they had beforehandfashioned or were allocated. These schemes are disposed toequitation from either a brute force effort to blow the passwordor from passwords which were not sole or even which weredispatched adjacent the computer itself. A BiometricIdentification scheme is one in which the user's "body" turn outto be the password/PIN. Biometric features about the distinctare what make that person sole and consequently can be cast-off to validate a user's admittance to numerous schemes.

Biometrics bids sole assistances for recognizing human beings.Tokens, Photo ID cards can be misplaced, pinched, replicatedor left at home, Passwords can be elapsed communal or

perceived but Biometrics grips the potential of firm, easy-to-use, perfect, consistent, and less costly authentication for anassortment of applications. The biometric identificationmethods substitute the token recognition scheme where aperson is proved by means of his license or pass codes andfavored over outdated passwords and PIN-based approaches.Certain recognized Biometric identification methods are statedunderneath, “Fingerprint recognition, Voice/sound recognition,Palm vein recognition, Iris recognition, Retina scan recognitionand so on”.

Denoising is a crucial issue in image process. Impulse noisesarea unit evoked as a result of bad pixels in camera sensors orbroadcast during a clattering channel. Two common forms ofimpulse noise area unit the random valued noise and the salt-and-pepper noise. For the photographs tainted by salt-and-pepper noise, clattering pixels take solely the utmost and theleast standards within the dynamic vary. These noises will

ISSN 0970-938Xwww.biomedres.info

Biomedical Research 2018; Special Issue: S31-S39

S31Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 2: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

scale back the standard of pictures and harm the appearance offacts for pictures effectively. Image filtering will effectivelyscale back the noise within the image. The objective of impulsenoise elimination is to subdue the noise to realm thetruthfulness of edge and aspect info. There are a unit severalefforts on the refurbishment of pictures tainted by impulsenoise. An upright and well-organized noise removal methodought to take away most noise further because it ought to retainthe necessary feature the maximum amount as potential. Thesame old strategies of image filtering normally are spatialfiltering and frequency domain filtering. There are numerousfiltering strategies accessible to de-noise the image. Earliertechniques notice the neighboring extreme “noisy” pel to befiltered out. Alas, the intense pixels that area unit considerednoise may additionally contain the first details that may furtherbe removed throughout the smoothing method.

An economical denoising technique is projected victimizationreflected CA supported Moore neighborhood (8- neighborhoodcell), Iterative Refined clattery picture element Restoration(IRNPR) with mirrored CA. This image denoising method isexamined with numerous prevailing image denoising methodslike, Median filter, Switch Median (SM) filter, DirectionalWeighted Median (DWM) filter, changed DWM (MDWM),Fuzzy Cellular Automata (FCA), changed call based mostlyUnsymmetric cut Median Filter (MDBUTMF), and Contramean (CHM) filter. 3 sample pictures (Finger print, Iris, andMandrill) of 2 completely different resolutions (512 × 512) and(256 × 256) area unit taken for the performance analysis.IRNPR-mirrored CA is examined in contradiction of PeakSignal-to-Noise Ratio (PSNR), Mean square Error (MSE), andStructural SIMilarity (SSIM). It’s determined that projectedmethod performs higher than the opposite existing technique interms of PSNR. The remainder of this effort is prearranged aslower than: Works that square measure associated with thedenoising of pictures square measure given in Section II Theoperating and outline of the planned method for the noiseconstituent recognition and reinstating of gray scale picturessquare measure given in an elaborate way in section III.Potency and recital of the planned method square measureexamined over investigating within the section IV. Finally,section V concludes with a thought to reinforce the denoisingtechnique within the future [1,2].

Related WorkProtter and Elad [3] have used redundant and distributeddepictions for eliminating noise from the image. Associate innursing algorithmic rule named K-SVD was accustomed train asparsified lexicon for the pictures that were corrupted. Here,the students generalized the higher than mentioned algorithmicrule through the subsequent ways in which (i) range ofrepetitions were abridged for broadcasting the lexicon fromone border to a different or subsequent frame and (ii) coverswere be close to in each temporal and spatial adjoiningpositions. These mentioned ways in which were accustomedhave a substantial profit in complexness and denoisingperformance. A distinctive method termed support vector

regression (SVR) was functional by the investigators of [4] soas to acquire purged of noise from a picture. The supportvector standards and their weights were calculated when thevociferous pictures were qualified with a ground-truth. Thesecalculated values were accustomed take away the haphazardnoise gift in a picture at varied stages on a pixel-by-pixel root.This can be Associate in nursing example-based approach as aresult of its cast-off SVs for noise elimination. Theinvestigational outcome study bestowed during this objectexhibited that the SVR primarily based denoising methodologyperforms higher than the Besov ball forecast method on theimage that was non-natural by means of each PSNR and visualexamination.

Researchers of the work [5] have cast-off PCA in conjunctionwith native constituent combination (LPG) for noiseelimination. With the intention to realm the nativeconfigurations, a vector variable was sculptural from aconstituent and their adjacent neighbor. Their examples forcoaching were non-appointive from the native window overblock matching primarily based LPG. This methodologyensures that solely alike at ease and example blocks were cast-off for the approximation of PCA alteration. For higherperformance, LPG-PCA methodology was iterated.Investigational outcome specified within the paper uttered thatthe LPG-PCA methodology outperforms the progressive noiseelimination procedures. Another technique was projected in [6]for removal of random noise. This methodology used Nonlocalsuggests that formula for economical noise elimination andalso the outcomes of the trial exhibited that NL-meansprimarily based formula accomplished than the progressivedenoising formula. Changhong et al. [7] proposes a completelyunique enhanced median filter formula for the photographsextremely tainted with salt-and-pepper noise. Primarily,altogether the pixels square measure classified supported thenative datum data into signal pixels and crying pixels bymistreatment the Max-Min noise detector.

The crying pixels square measure distinguished into 3categories like small, modest, and great-density noises. Finallythe biased 8-neighborhood resemblance perform filter, the five× five median filter and also the 4-neighborhood mean filtersquare measure adopted to get rid of the noises. The papers [8]define the assiduousness of cellular automata (CA) for manyimage process tasks like denoising and have detection. Theaccrued range of cell states (i.e. constituent intensities) resultsin an enormous increment within the range of potential rules.Consequently, an abridged intensity illustration was used,resulting in a 3 state cellular mechanisms that was additionalsensible. In concurrence, a changed ordered floating forwardsearch technique was developed so as to hurry up the selectionof fine rule sets within the CA coaching stage. In paper [9] areplacement methodology was projected to eradicate noise andto determine image edges through then perform of fuzzycellular automata as in [10]. During this methodology, eightspecific contiguous circumstances square measure consideredfor every component and sixteen numbers area unit resultingfrom these transitions. Such numbers area unit cast-off as effortfor the fuzzy membership performs. The fuzzy rule ground was

Suresh/Malathi/Nagarani/Oswalt

S32Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 3: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

made in such the simplest way on properly acknowledge thetransition of every component.

A changed call based mostly Unsymmetrical cut Median Filter(MDBUTMF) joined with Fuzzy Noise Reductionmethodology (FNRM) was planned [11] for the refurbishmentof color pictures that are tremendously tainted by salt andpepper noise. This filter alternates the howling picture elementby cut norm once the weather with 0’s and 255’s values aregift. The partial noise detached pictures are additional treatedby FNRM to get a completely noise unconcerned imageassociate amended form of the directional weighted median(DWM) filter was planned in paper [12]. This methodologywill considerably progress the recital of the directional-weighted-median filter because it includes further ways (12directions) for edge detection, while the DWM filter solelypays four directions. These further guidelines amend theprecision of edge recognition.

Esakkirajan et al. [13] detached the greater thickness salt andpepper noise in grey scale and color pictures employing achanged call based mostly asymmetric cut median filter. Thisalgorithmic rule exchanges the howling picture element withthe cut norm, once all the picture element values within thedesignated window are 0’s and 255’s. The noise pictureelement was substituted by the average of all the weatherwithin the nominative window. Gorsevski et al. [14] perceivedgrain boundaries in misshapen stuns by cellular automatarefrain. Two-dimensional CA was functional to the removal ofskinny segments from misshapen rocks and grain boundaryrecognition. The haul out boundaries comprises options likealignment, shape, and spatial dissemination, shaped from a CAMoore’s environment-grounded instructions practice. TheMoore’s atmosphere entails of a three × three matrix of variedthe appearance by distinguishing amongst a middle pictureelement and its neighbors. These rules calculable the longerterm state of every cell whereas the amount of recurrences tofake boundary recognition was user-specified. The yield atevery spherical provides numerous recognition circumstancesof grain boundaries. Kumar and Sahoo [15] planned a uniquepractice for edge recognition exploiting CA.

Chhabria and Shende [16] intended a CA image removalalgorithmic rule for head and hand gestures acknowledgment.These real time vision systems are often applied in associatemultimedia atmosphere. It entails of image capture, imageremoval, pattern connotation, and command approximation.The gestures are joined to the pre-stored information ofgestures. Then, the hardware association was achieved as perthe joined signs on quad-directions. This movement also can becontrolled mistreatment voice commands. The cellularautomata accommodate a multi-dimensional array/grid of cells.These cells will inhabit several finite varieties of attainablestates. The presence of the cells are at the same time updated inline with a state transition performs. Aydogan [17] developed aCellular Neural Network (CNN) grounded mostly edgerecognition of second information. This CNN replica was cast-off for the recognition of the image body and edges. A randomimage process algorithmic rule was planned supported the

instantaneous vicinity link of the cells. This methodology wascast-off for edge recognition and image enhancement. CNNwas functional to arena and artificial information made foredge recognition of skinny surface earth science entities thatdefend one another in numerous depths and dimensions. Abiological process edge detector was projected victimizationvaried instruction CA by Sato and Kanoh [18]. The trainingformula cast-off here was organic phenomenon programming,wherever the transition instruction was determined because themanifestation by the body coaching CA was produced by smartedge detector and 2 impartial tasks were intended. The targetjobs were enhanced by a multi-objective biological processformula.

Qadir et al. [19] projected a completely unique image noisefiltering formula supported CA. This technique may also takeaway the impulsive noise from the rip-roaring pictures. Non-uniform CA rules were designed to get rid of the noise fromgeneral and medical pictures. Hsu, et al. [20] developed aformula for salt and pepper noise reduction methodvictimization CA. The software system programming modelreferred to as CAID (Cellular Automata Image Denoising)toolkit was established victimization MATLAB. The propertiesconcerning instinct or salt and pepper noise were stated bySelvapeter and Hordijk [21]. In impulse noise, solely ahaphazard part of image pixels are corrupted. These rip-roaringpixels whichever take terribly massive price (gray {scale price|value} 255) as an optimistic impulse or terribly little value zeroas undesirable impulse. Fastened price boundary conditions arapplied within the cellular automata filter in salt and peppernoise elimination, i.e., update instruction is functional solely tothe non-boundary cells. The bulk CA update instruction wontto take away impulse noise was mentioned within the paper.The noise propagation stemmed from the settled CA rule isresolved by employing a haphazard instruction for breakingmainstream ties. A CA replica for eliminating salt and peppernoise and uniform noise was projected by Dalhoum et al. [22].The projected CA model at the start checks the kind of noiseby calculating the bar graph of the rip-roaring image. The CAtransition instruction that is employed to get rid of the salt andpepper noise was mentioned within the paper. The methods forcoaching CAs to implement justly common place imageprocess responsibilities to a great level of recital for projectedby Rosin [23]. In CA, the fastened price boundarycircumstances are functional during which transition states areenforced solely to non-boundary cells. The instruction set isverified at every pel to confirm if any of its instructionsapproves the pel neighborhood pattern. The paper designatesthat the method of applying the principles are continued till theamount of iterations has reached a planned most.

Image Denoising Using Mirrored CAThis segment defines the image denoising method exploitationIRNPR-mirrored CA for the biometric recognition system.First, the summary of the noise filtering methodology isinstructed with the design. Then, the IRNPR-mirrored CA forincreased image denoising is mentioned.

An improved cellular automata (ca) based image denoising method for biometric applications

S33Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 4: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

CA and its characteristicsBased on the adjoining cell’s states, the CA replica transitioninstruction decides the neighborhood association amongst theautomata. The formal automaton at time (t) be influenced byon the state of adjacent cells at time (t-1). There are a unit fourcategories to classify all potential behavior in CA. In class 3,the majority initial configurations relax once a transientamount to apparently unpredictable area time behavior. Theuniform cellular automata area unit category three.

Application of CA in image processUniform cellular automata instructions area unit engineered toget rid of impulse blare from binary and grey scale pictures. Ahaphazard CA instruction is employed to reduce noisepropagation gift in settled CA filters. A Moore neighborhood(the eight adjoining cells encompassing a cell) is taken intoaccount. A 2-D CA with an easy modernize rule is functionalas effective impulse noise filter in digital pictures. CAimplements mounted price boundary conditions wherever thefill in rule is functional solely to non-boundary cells. Thepreliminary CA lattice outline is that the image tainted byAssociate in Nursing impulse image. The mainstream CAmodernize rule tells that if the middle constituent grey level iszero or 255, then the bulk grey level within the nativeneighborhood is employed to interchange the middle pixel’sprice. The settled or random CA philosophy is employed oncethere's no majority grey level within the native neighborhood.Within the random philosophy, the grey price ofindiscriminately chosen constituent within the neighborhoodreplaces the middle constituent. Within the settled philosophy,the mounted grey level within the neighborhood replaces themiddle constituent. The noise broadcast occasioned from thesettled CA rule is resolved by haphazard rule. A random CAperforms higher than the settled CA for grey level pictures. CAtransition rule is employed within the elimination of salt andpepper noise. This rule drafts if the present constituent iscorrupted. Just in case of prevalence of noise, if all adjacentarea unit clamorous, the present constituent is substituted bythe typical of neighbors. If all neighbors aren't clamorous, thentheir median replaces the present automation state. Themounted boundary downside is solved by reflected CA.

Design of noise filteringPrimarily, a check image is administered to notice theoccurrence of impulse noise. Any constituent image is chosenand patterned either or not it's noise free or clamorous. Thiscan be done by checking whether chosen constituent price ismost or least (0 or 255). If the worth is either zero or 255 thenthe constituent is clamorous and managed by repetitiouspurification technique. Else the values is not zero or 255, theconstituent is noise open and remains unaffected.

Where In the salt and pepper noise, the tainted constituenttakes one in all the 2 totally different values: black or white.The median sieve is employed to get rid of unsolicited noiseand upsurge image excellence. CA entails of interconnected

cells, every of that has Associate in nursing automaton.Associate in Nursing automaton that may be a mechanicaldevice wont to execute easy calculations encompasses a stateever-changing.

Added, a three × three window focused at (i, j) is functional tothe clamorous constituent. to interchange this clamorousconstituent, the neighbor constituents of this pixel area unitthought-about exploitation IRNPR-mirrored CA, and sochecked whether or not all the neighboring pixels area unitclamorous or not. Paste your text here and click on “Next” tolook at this text redactor do its issue. Don’t have any text tocheck? Don’t have any text t check? Click “Select Samples”.��, � = 0   ��   255       �����   ����������ℎ���     �����   ���� (1)Where Xi,j = Intensity value of pixel at location (i,j)��, � = �����     ,��������   �����  ����   �������   ����   ,��   �ℎ���� (2)Further, a three × three window targeted at (i, j) is applied tothe clanging picture element. to exchange this clanging pictureelement, the neighbor picture elements of that pixel at thought-about exploitation IRNPR-mirrored CA, and so checkedwhether or not all the adjoining pixels are clanging or not. CAwith buffering scheme is employed within the plannedmethodology. i.e., whereas hard the worth of elite pictureelement at time t, the worth of adjoining pixels at time (t-1)hold on within the buffer is employed. CA may be outlined as(I, N, V, F). I could be a cellular house fashioned by a 2dimensional array of maximum cells, I = {(a,b),1 ≤ a ≤ m, 1 ≤b ≤ n}, N – form of adjacency (Moore neighborhood), V – Setof adjacency picture element price, F- Transition Perform. If alladjacent pixels are clanging i.e. they need values either zero or255 as shown in (3), then they're thought-about as clangingpicture element and also the window is extended to 5 × 5.255 0 2550 〈 255 〉 2550 255 0 (3)If a minimum of a number of the pixels don't seem to berackety, then the median of the neighbor pixels are calculatedby excluding the rackety pixels (0 or 255). The chosen pictureelement is substituted with this median. Then the method istouched to next picture element. Deliberate a 3 × 3 window asshown in fig two. Here X(i, j) is that the chosen pictureelement and alternative pixels are subsequent of X(i, j).��, � = ∀� �, �   ���   �����   ������   ��  �5 × 5��ℎ������   �   ���������     0   ��   255 (4)Where p=comprises 8 neighbor pixels of X(i,j), If the noiseperseveres in 5x5 window additionally, the window isprolonged to 7 × 7. If the pixels aren't blatant notice the normof residual pixels of 5 × 5 window and reinstate the worth toX(i, j) as in equation five. If all the adjacent pixels ar blatant in7 × 7 window, the norm of adjacency is considered else if anumber of the adjacent pixels aren't blatant in 7 × 7 window,

Suresh/Malathi/Nagarani/Oswalt

S34Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 5: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

then the norm of remaining pixels without blatant pixels ispremeditated. Then the chosen picture element is substitutedwith the intended worth.��, � = ∀� �, �   ���   �����   ,   ����   ��   ���   � �, ���ℎ������   �   ���������     0   ��   255(5)��, � = ∀� �, �   ���   �����   , ������   �����∀� �, �   ���������     , ��� (6)��� = ���   ���   , ������ℎ   ���   ��� (7)Finally, once using these procedures for each pel within theimage, the reinstated image is attained. The standard of theimage is examined victimization Mean Square Error (MSE). Ifthe MSE of the de-noised image is below the MSE ofpreceding generation cellular automata, the image is noisepermitted. Or else the image entails of noise. Once anoccasional MSE reinstated image is attained, that proceduresthe ultimate de-noised noise- open image.

Performance AnalysisThe planned image denoising methodology is analyzed for fourtotally {different|completely different} grey scale pictures of 2different resolutions every. The pictures square measure fingerprint (512 × 512), (256 × 256), Iris (512 × 512), (256 × 256)model of unique Finger print and Iris pictures square measurespecified in Figure 1.

Figure 1. Sample images (Left to Right): Finger print and Iris.

Table 1. MSE, PSNR and SSIM values for 10% salt and Pepper noisein Finger print image.

Methods MSE PSNR SSIM

Noisy Image 1.8605e+03 15.4344 0.4019

Without CA 27.2664 33.7745 0.9784

With CA 3.1119 43.2006 0.9977

Outcomes for 10% salt and pepper noiseThe output results of denoised image sample to fingerprint isshown in Figure 2 (without CA)

Figure 2. Image finger print (Left to Right): Input noisy image anddenoised image (without CA).

Figure 3. Image finger print (left to Right ) Noisy Image and denoisedimage (with CA).

The output results of denoised image sample to fingerprint isshown in Figure 3 (with CA) , Table 1 shows the performancecomparison results of the denosiy images related to fingerprintwith three different parameters (MSE,PSNR and SSIM).

Results for 90% salt and pepper noiseThe output results of denoised image sample (salt 90% andpepper noise) to fingerprint as shown in Figure 4 (without CA).The output results of denoised image sample (salt 90% andpepper noise) to fingerprint is shown in Figure 5 (with CA).

Figure 4. Image Finger print (Right to Left): Noisy input image anddenoised image (without CA).

Figure 5. Image Finger print (Left to Right):Noisy image anddenoised image(with CA).

An improved cellular automata (ca) based image denoising method for biometric applications

S35Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 6: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

Table 2 shows the performance comparison results of thedenosiy images related to fingerprint with three differentparameters (MSE, PSNR and SSIM).

Table 2. MSE, PSNR and SSIM values for 90% salt and Pepper noisein Finger print image.

Methods MSE PSNR SSIM

Noisy image 1.6701e+04 5.9034 0.0273

Without CA 1.4083e+04 6.6437 0.0247

With CA 172.8126 25.7550 0.8682

Iris image – 10% noiseThe output results of denoised image sample (salt 10%) to irisare shown in Figure 6 (without CA). The output results ofdenoised image sample (salt 10%) to iris are shown in Figure 7(with CA).

Figure 6. Image Iris (Left to Right): Noisy input image and denoisedimage (without CA).

Figure 7. Image Iris (Left to Right): Noisy image and denoised image(with CA).

Table 3 shows the performance comparison results of thedenosiy images related to iris (10% noises) with three differentparameters (MSE, PSNR and SSIM).

Table 3. MSE, PSNR and SSIM values for 10% noise in iris image.

Methods MSE PSNR SSIM

Noisy Image 1.8589e+03 15.4382 0.4576

Without CA 69.4081 29.7167 0.9581

With CA 8.0039 39.0978 0.9952

Iris–90% noiseThe output results of denoised image sample (salt 90%) to irisare as shown in Figure 8 (without CA). The output results ofdenoised image sample (salt 90%) to iris are shown in Figure 9(with CA).

Figure 8. Image Iris (Left to Right): Noisy input image and denoisedimage (without CA).

Table 4 shows the performance comparison results of thedenosiy images related to iris (90% noises) with three differentparameters (MSE, PSNR and SSIM).

Figure 9. Image Iris (Left to Right): Noisy image and denoised image(with CA).

Table 4. MSE, PSNR and SSIM Values for 90%noise in iris image.

Methods MSE PSNR SSIM

Noisy Image 1.66E+04 5.9329 0.0317

Without CA 1.40E+04 6.6607 0.0284

With CA 304.2348 23.2987 0.7909

Contrast with prevailing waysThe images are examined by means of Peak Signal-to-NoiseRatio (PSNR) [24], Mean Squared Error (MSE) [25], andStructural SIMilarity (SSIM) [26].

��� = ∑�∑� � �,� − � �,� 2� × � (8)����  ��   =  10 ���10 2552/��� (9)���� = 2����+ �1 2���+ �2��2 + ��2 + �1 ��2 + ��2 + �2 (10)

Suresh/Malathi/Nagarani/Oswalt

S36Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 7: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

Two non-negative image signals a and b are associated withone another for the dimension of SSIM [26]. The subsequentsymbolizations are cast-off in equation 10 (Table 5).

Table 5. PSNR (dB) for Finger print (512 × 512) image for diversewindow sizes, numerous noise densities and IRNPR-mirrored CA.

Density 3 × 3 3 × 3 or 5 × 5 Refined 3 × 3

or 5 × 5Iterative refined 3 × 3or 5 × 5 or 7 × 7

10 42.97 42.91 42.9944 43.0698

20 39.20 39.14 39.1135 39.2165

30 36.64 36.61 36.7338 36.6313

40 33.87 34.56 34.9466 34.8904

50 29.98 32.63 33.3527 33.1090

60 24.49 29.14 31.6570 31.8089

70 19.36 23.33 29.9900 30.1723

80 14.26 16.66 26.0760 28.4337

90 9.76 10.72 16.8664 25.9115

PSNR is examined for Finger print (512 × 512) image fordiverse window sizes, several noise densities and IRNPR-mirrored CA. The comparative studies are specified in Table 6.

Table 6. PSNR for Finger print image (512 × 512) by means ofIRNPR-mirrored CA and other Prevailing techniques.

Density SM DWM MDWM FCA CHM IRNPR- mirroredCA

10 36.12 40.78 41.45 35 37.31 43.06

20 33.42 37.02 38.22 33.50 34.59 39.21

30 31.36 34.63 35.97 32.70 32.99 36.63

40 29.88 32.51 34.07 31.20 33.01 34.89

50 28.54 30.23 32.69 29.20 29.53 33.10

60 26.76 27.69 31.21 27.80 27.20 31.80

70 24.47 25.23 29.72 26.50 26.45 30.17

80 19.52 21 27.94 23.70 22.56 28.43

90 8.80 15.45 25.50 18.37 19.57 25.91

The subsequent notations are cast-off in equation 10:

• μa is the average of a.• μb is the average of b.• σa

2 is the variance of a.• σb

2 is the variance of b.• σab is the covariance of ab.• C1 and C2 are two constants to evade uncertainty• [26].

Figure 10. Graphical exploration of PSNR for 2 two resolutions ofFinger print image (512 × 512) and (256 × 256), by IRNPR-mirroredCA and other obtainable procedures.

Figure 11. Graphical exploration of PSNR for Finger print image(512 × 512) by IRNPR-mirrored CA and other prevailingapproaches.

Figure 12. Graphical analysis of PSNR for Finger print (256 × 256)image using IRNPR-mirrored CA and other existing methods.

The prevailing image denoising techniques taken for theexamination are Median filter, Switch Median (SM) filter [1],Directional-Weighted-Median (DWM) filter [12], ModifiedDWM (MDWM) [4], Fuzzy Cellular Automata (FCA) [10],Modified Decision Based Unsymmetric Trimmed MedianFilter (MDBUTMF) [19], and Contra Harmonic Mean (CHM)filter [10]. PSNR is evaluated for two resolutions of the Fingerprint image (512 × 512), and (256 × 256), with IRNPR-mirrored CA and other prevailing approaches. The comparativeexploration is specified in Figure 10. PSNR is evaluated for

An improved cellular automata (ca) based image denoising method for biometric applications

S37Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 8: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

Finger print image (256 × 256) with IRNPR-mirrored CA andother prevailing approaches. The comparative studies arespecified in Figures 10 and 11. PSNR, MSE, and SSIM areexamined for Finger print (512 × 512) image for diverse noisedensities by means of IRNPR-mirrored CA (Figure 12). Therelative examination is specified in Table 7.

Table 7. PSNR, MSE, SSIM for Finger print (512 × 512) image fordiverse noise densities by means of IRNPR-mirrored CA.

Noise Density PSNR(dB) MSE SSIM

10 43.06 3.3 0.99

20 39.21 7.78 0.99

30 36.63 14.12 0.99

40 34.89 21.08 0.98

50 33.1 31.78 0.97

60 31.8 42.87 0.96

70 30.17 62.49 0.95

80 28.43 93.26 0.92

90 25.91 166.69 0.86

Conclusion and Future WorkAn economical denoising technique is projected exploitationreflected CA supported Moore neighborhood (8-neighborhoodcell). This image denoising technique is analyzed with manyexisting image denoising techniques like, Median filter, SwitchMedian (SM) filter, Directional-Weighted-Median (DWM)filter, changed DWM (MDWM), Fuzzy Cellular Automata(FCA), changed call primarily based asymmetrical cut MedianFilter (MDBUTMF), and Contra mean (CHM) filter. 3 samplepictures (Finger print, Iris, and Mandrill) of 2 totally differentresolutions (512 x 512) and (256 x 256) are taken for the recitalexamination. The reflected CA is analyzed against Peak ratio(PSNR), Mean square Error (MSE), and Structural SIMilarity(SSIM). It’s ascertained that reflected CA performs higher thanthe opposite existing methodologies in terms of PSNR. Thisreflected CA primarily grounded image denoising method isdistended to greater resolution image process with quickercomputing methods such as parallel computing.

References1. Toh KKV, Isa NAM. Noise Adaptive Fuzzy Switching

Median Filter for Salt-and-Pepper Noise Reduction. SignalProcess Lett IEEE 2010; 17: 281-284.

2. Dharmarajan R, Kannan K. A hypergraph-based algorithmfor image restoration from salt and pepper noise. AEU Int JElectron Commun 2010; 64: 1114-1122.

3. Protter M, Elad M. Image Sequence Denoising via Sparseand Redundant Representations. IEEE Transact ImageProcess 2009; 18: 27-35.

4. Li D. Support vector regression based image denoising.Image Vision Comput 2009; 27: 623-627.

5. Zhang L, Dong W, Zhang D, Shi G. Two-stage imagedenoising by principal component analysis with local pixelgrouping. Pattern Recogn 2010; 43: 1531-1549.

6. Lu K, He N, Li L. Nonlocal Means-Based Denoising forMedical Images. Comput Math Methods Med 2012; 2012:1-8.

7. Changhong W, Taoyi C, Zhenshen Q. A novel improvedmedian filter for salt-and-pepper noise from highlycorrupted images. Syst Control Aeronaut Astronaut(ISSCAA) 2010.

8. Rosin PL. Image processing using 3-state cellular automata.Computer Vision Image Understanding 2010; 114:790-802.

9. Mirzaei K, Motameni H, Enayatifar R. New method foredge detection and de noising via fuzzy cellular automata.Int J Phy Sci 2011; 6: 3175-3180.

10. Sadeghi S, Rezvanian A, Kamrani E. An efficient methodfor impulse noise reduction from images using fuzzycellular automata. Ommunications 2012; 66: 772-779.

11. Vimala T. Salt and pepper noise reduction using mdb utmfilter with fuzzy based refinement. Int J Management ITEng 2012; 2: 447-461.

12. Lu CT, Chou TC. Denoising of salt-and-pepper noisecorrupted image using modified directional-weighted-median filter. Pattern Recogn Lett 2012; 33: 1287-1295.

13. Esakkirajan S, Veerakumar T, Subramanyam AN,PremChand CH. Removal of High Density Salt and PepperNoise Through Modified Decision Based UnsymmetricTrimmed Median Filter. IEEE Signal Process Lett 2011; 18:287-290.

14. Gorsevski PV, Onasch CM, Farver JR, Ye X. Detectinggrain boundaries in deformed rocks using a cellularautomata approach. Computers Geosci 2012; 42: 136-142.

15. Kumar T, Sahoo G. A Novel method of edge detectionusing cellular automata. Int J Comput Appl 2010; 9: 38-44.

16. Chhabria SA, Shende RS. Cellular Automata ImageExtraction Algorithm for Gestures Recognition. Int JComput Sci Technol 2011; 2: 251-255.

17. Aydogan D. CNNEDGEPOT: CNN based edge detection of2D near surface potential field data. Comput Geosci 2012;46: 1-8.

18. Sato S, Kanoh H. Evolutionary design of edge detectorusing rule-changing Cellular automata. Nature BiolInspired Comput (NaBIC) 2010.

19. Qadir F, Peer MA, Khan KA. An effective image noisefiltering algorithm using cellular automata. ComputCommun Informatics (ICCCI) 2012.

20. Hsu CY, Tsui TS, Yu SS, Tseng KK. Salt and Pepper NoiseReduction by Cellular Automata. Int J Appl Sci Eng 2011;9: 2-19.

21. Selvapeter PJ, Hordijk W. Cellular automata for imagenoise filtering. Nature Biol Inspired Comput 2009; 2009:193-197.

22. Sakthivel K, Jayanthiladevi A, Kavitha C. Automaticdetection of lung cancer nodules by employing intelligent

Suresh/Malathi/Nagarani/Oswalt

S38Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue

Page 9: An improved cellular automata (ca) based image denoising ...€¦ · An improved cellular automata (ca) based image denoising method for biometric applications. Suresh A 1*, Malathi

fuzzy cmeans and support vector machine. Biomed Res2016; 27: S123-S127.

23. Rosin PL. Training Cellular Automata for ImageProcessing. IEEE Transact Image Process 2006; 15: 1-12.

24. Esakkirajan S, Veerakumar T, Subramanyam AN,PremChand CH. Removal of High Density Salt and PepperNoise Through Modified Decision Based UnsymmetricTrimmed Median Filter. Signal Process Lett IEEE 2011; 18:287-290.

25. Zhou W, Bovik AC, Sheikh HR, Simoncelli EP. Imagequality assessment: from error visibility to structuralsimilarity. Image Process 2004; 13: 600-612.

26. Hsieh MH, Cheng FC, Shie MC, Ruan SJ. Fast andefficient median filter for removing 1–99% levels of salt-and-pepper noise in images. Eng Appl Artificial Intell2013; 26: 1333-1338.

*Correspondence toSuresh A

Department of Information Technology

Sri Ramakrishna Institute of Technology

India

An improved cellular automata (ca) based image denoising method for biometric applications

S39Special Section: Computational Life Sciences and Smarter Technological Advancement

Biomed Res 2018 Special Issue