Parameter Controlled by Contrast Enhancement Using Color Image

Parameter Controlled by Contrast Enhancement Using Color Image Raguathi.S and Santhi.K Abstract -The arameter-controlled virtual histogram distribution (PCVHD) method is roosed in this roject to enhance the fusion image. This introduces a new hybrid image enhancement aroach driven by both global and local rocesses on luminance and chrominance comonents of the image simultaneously the overall contrast and the sharness of an image also increased. An original image creates equally distributed brightness level. This aroach also increases the visibility of secified ortions or better maintaining image colour. Primary oeration for almost all vision and image rocessing tass in several areas such as comuter vision, biomedical image analysis, forensic video/image analysis, remote sensing and fault detection is to enhance the visual information. A simulation result shows better brightness reservation. Index terms Image Enhancement, Contrast Enhancement, Brightness reservation, Histogram Equalization. I. INTRODUCTION In most of Image enhancement which transforms digital images to enhance the visual information within is a rimary oeration for almost all vision and image rocessing tass in several areas such as comuter vision, biomedical image analysis, forensic video or image analysis, and remote sensing [2, 4]. For examle in forensic video/image analysis tass surveillance videos have quite different qualities comared with other videos such as the videos for high quality entertainment or TV broadcasting. Enhancement transformation to modify the contrast of an image within a dislay s dynamic range is therefore required in order to reveal full information contents in the videos. The most requirements for image rocessing of images are that the images be available in digitized form that is arrays of finite length binary format. For digitization the given image is samled on a discrete grid and each samle or ixel is quantized using a finite number of bits. The digitized image is used in comuters. S.Raguathi, PG Student, K.S.Rangasamy College of Technology, Tiruchengode. INDIA (+91 9952666407); A digital image it is first converted into analogy signal which is scanned into a dislay [7, 9, 11].Before going to rocessing an image it is converted into a digital format. Digitization includes samling of image and quantization of samled values. After converting the image into bit information rocessing is erformed. Surveillance videos have quite different qualities comared with other videos such as the videos for high quality entertainment or TV broadcasting. High quality entertainment movie or game videos are roduced under controlled lighting environment where as surveillance videos for monitoring outdoor scenes are acquired [4, 8].Recently develoed more Histogram aroaches based on Dynamic Range Searate Histogram Equalization (DRSHE)[4], Minimum Mean Brightness Error Bi- Histogram Equalization(MMBEHE)[6] and Brightness Preserving Dynamic Histogram Equalization (BPDHE)[3] develoed in order to overcome of the some Histogram Equalization methods. Most one of the common defects of surveillance videos is oor contrast resulting from reduced image chrominance. This aroach driven by both global and local rocesses of fusion image comonent to enhance the visual information the maintaining image colour Primary oeration for almost all vision and image rocessing tass in several areas such as comuter vision, biomedical image analysis, forensic video or image analysis, remote sensing. The rest of organized as follows; Section II the roosed methodology is discussed. Section III resents the results and discussion for the roosed wor followed by conclusion. II. PRINCIPLE OF THE PROPOSED METHOD In the erformance evaluation of the roosed method which wors as an automatic enhancement method using arameters with default values is comared with four classical enhancement methods (gamma correction, histogram equalization, linear contrast stretching and Virtual Enhanced color outut).[2, 6, 9] and some recent develoed histogram equalization based methods. K.Santhi, Assistant Professor, K.S.Rangasamy College of Technology,Tiruchengode. INDIA (+91 9843281823); 558

A.Flow Diagram Inut Image Gamma Correction Histogram Equalization Contrast Stretching Virtual Enhanced This curve ixel values range from 0.0 reresenting ure blac to 10.0 which reresents ure white. Fig 2 shows that the examle for Gamma values of less than 1.0 daren an image. Gamma values greater than 1.0 lighten an image and a gamma equal to 1.0 roduces no effect on an image. While maintaining the dynamic range of the image fact that gamma Correction introduces secific higher-order correlations in the frequency domain [1]. B. Histogram Equalization The Histogram is a grahical reresentation showing a visual imression of the distribution of data. Image histograms are resent on many modern digital cameras are available. Photograhers can show the distribution of tones catured and whether image detail has been lost to blown-out highlights or blaced-out shadows. Conversely the histogram for a very bright image with few dar areas and/or shadows will have most of its data oints on the right side and center of the grah [3, 5, 7]. Fig 1. Flow diagram of Virtual Histogram Aroach. A significant amount of tested color images better results of the comared techniques, such as Gamma Correction, Histogram Equalization, Contrast Stretching and Virtual Enhanced Methods are obtained by the entire flow diagram shows the Fig 1. 1) Gamma Correction Gamma correction is adjusting the lightness/darness level of their rints. This method using ower law s light intensity oeration [1]. The luminance non-linearity introduced by many imaging lighting devices can often be described with a simle oint-wise oeration. (1/gamma value) X=(x) Where x is the original ixel value and gamma value1.0. thus, the overall image can be lightened or darened deending on the gamma value used. The amount of correction is secified by a single value ranging from 0.0 to 10.0. 0.0 1.0 10.0 Fig 2. Gamma s. Darer no change Lighter Comarison of arameter calculation For a given image X, the robability density function P(X ) is defined as P(X ) = P(X ) = n n number of ixels with intensity n Total numberof ixels For = 0, 1,, L 1, (1.1) Where, n reresents the number of times that the level X aears in the inut image X and n is the total number of samles in the inut image. HE is a scheme that mas the inut image into the entire dynamic range, (0, L-1), by using the robability density function as a transform function. Let s define a transform function F(x) based on the robability density functions, F(x) = (L-1) x (x)dx 0 (1.2) Where, P(x) is the robability density function of F(x). The robability distributive function of 0 to L 1. L is the number of ossible intensity values often 256. 1) Bi-Histogram Equalization Brightness Preserving Bi-Histogram Equalization based on these transform functions the decomosed sub-image constitute the outut of BBHE.The robability density function of sub-images i and j is defined as P i (I ) = n i P j (I ) = n j n i (1.3) n j (1.4) Where, n i and n j reresent the resective values of I in the two sub-images i and j, then n i and n j are the total values of i and j resectively. The mathematically resective Cumulative Density Function (CDF) are then defined as 559

P i (I ) = P j I = i=im i I (1.5) j =I m +1 j I (1.6) Where P i ( I ) =1 and P j I = 1 by definition. The transformation functions exloiting the CDFs. Recently some histogram based aroaches have been develoed in order to overcome some drawbacs of histogram equalization methods [3,5,7] such as, Dynamic Range Searate Histogram Equalization (DRSHE). Brightness Preserving Dynamic Histogram Equalization (BPDHE). Gain-Controllable Clied Histogram Equalization (GC-CHE). C. Contrast Stretching In the method the histogram of the enhanced colour image should not be saturated at one or both ends of the dynamic range or at least not bring new significant sies at the tail ends in order not to introduce new oor exosure lie defects in the image [8, 10]. In addition to a full use of the maximum ossible dynamic range colour comonent and local information can certainly mae a contribution to the contrast enhancement. The YC B C R colour sace reresents the ixel multivariate value. Each RGB colour channel each individual histogram entry is defined as sometimes transformed from the RGB colour sace or another colour sace to the YC B C R colour sace necessarily in the roosed image enhancement [2, 8]. The cumulative histogram for each RGB comonent and luminance comonent Y for the YC B C R colour sace are defined by extending the definition of cumulative histogram from grey-scale image resectively as A simle image enhancement technique that attemts to imrove the contrast in an image by `stretching' the range of intensity values it contains to san a desired range of values [9].The stretching can be erformed it is necessary to secify the uer and lower ixel value limits over which the image is to be normalized. In Fig 3 shows the examle for 8-bit gray level images the lower and uer limits might be 0 and 255.Fig 1 mention that ranges is 0 to 20 lower thresholding levels and 230 to 255 ranges is uer thresholding level. 20 to 230 ranges are contrast stretching outut level. Similarly the intensity histogram is then scanned downward from 255 until the first intensity value with contents above the cut-off fraction. 1) Local and Global Contrast Stretching Local Contrast Stretching (LCS) is an enhancement method erformed on an image for locally adjusting each icture element value to imrove the visualization of structures in both Darest and lightest ortions of the image at the same time. 2) Partial Contrast Partial contrast is an auto scaling method. It is a linear maing function that is usually used to increase the contrast level and brightness level of the image. This technique will be based on the original brightness and contrast level of the images to do the adjustment. H R x, y = H G x, y = H B x, y = h R i (1.7) h G i (1.8) h B i (1.9) And H Y x, y = h Y i (2.0) Where the inut brightness value is [ o, ].The cumulative histograms are monotonic no-decreasing functions. Comared with the original image an enhanced image with good contrast will have a higher intensity of the edges. The inut brightness values are [ 0, ] find out this H, H 1 and H 2 are defined as follows, H x, y = H 1 x, y = w And H 2 x, y = v h Y i (2.1) 1 i= 1o h Yw i (2.2) h Yv i (2.3) Where and 10 are in the range of [ 0, ] [0,255]; w is a arameter with the default value set to 2. v is a Parameter with the default value set to 1. Here H 1 is designed to suit secial enhancement requirements for the image interretation. c n = H H+H 1 +H 2 (2.4) Where normalized coefficient a new virtual distribution function is defined h o i = c n ( h Y i Fig 3. Partial contrast Auto scalling method. D. Virtual Enhancement +w 1 i= 1o h Yw i v h Yv i ) (2.5) If M and N are the height and the width of an image resectively and the outut brightness range is [ 0, ].The desired outut histogram can be aroximated by its corresonding continuous robability density as follows 560

MN q 1 ds = h q q o (s)ds (2.6) q o o o The left side of Equation is the corresonding uniform robability distribution function. The desired ixel brightness histogram transformation T is defined as q = T = q q 0 MN h o s ds + q o o (2.7) The default values of these arameters are, q 0 = 0, = 255 and w = 2. The arameters can be adjusted by an image interreter meet to secific requirements [8]. For the various arameters wors well without user intervention as changes of the arameters do not affect the enhanced result very much show the results. In this case the roosed method can be alied directly to each of RGB channels using the contribution from the color and the luminance comonents for contrast enhancement. It shows clearly better erformance in some of the test images. (a) III. EXPERIMENTS RESULTS In the roosed method wor as analyzed with quality and quantitative of city light and boo house inut images. A significant amount of tested colour images better results of the comared techniques, such as Gamma Correction, Histogram Equalization, Contrast Stretching and Virtual Enhanced Methods are obtained by first converting the image to the Hue, Saturation, Intensity colour sace and then alying the comared techniques to the Intensity comonent only Images are catured from camera source will be tae a hoto image. Fig. 4 shows the inut color image results containing both dar and bright regions and result of the image enhancement results by Gamma Correction, HE, Contrast Stretching, and Virtual Enhanced roosed method. The original color of the image is changed after enhancing using HE and Virtual Proosed method as well as the contrast is increased necessarily. The test image resolution of 640x480 or 352x291 or 480x640 or 640x360 ixels. In all images the inclusion art has be enhanced and the arameter such as Mean Square Error (MSE), Pea Signal to Noise Ratio (PSNR), and Structural Similarity Index (SSI) is to be calculated. The Contrast Stretching failed to mae significant enhancement for these images shows the fig 4(d). Mean Square Error (MSE) value is deending uon the ixel resolution chances the brightness and contrast of image are low or high reserved as well as enhanced quantitative value are listed in Table I. Future wor will be segmentation of the inclusion from the alication image and it can otimize with hel of its shae and size. The original color of the image is changed after enhancing using HE and Virtual Proosed method as well as the contrast is increased necessarily. It shows clearly better erformance in some of the test images. (b) (d) Fig 4.Image Contrast Enhancement comarison with different method: (a) Boo house original inut image: remaining four images are the Enhancement results (b) Gamma Correction enhanced outut image. (c) HE enhanced outut image. (d) Contrast stretching enhanced outut image. (e) Virtual Enhanced outut image. (c) (e) (a) 561

V.REFERENCES (d) (b) Fig 5. (a) is the original image with corresonding statistical Histogram Plot: Remaining four images are modified histograms generated by the (b) Gamma Correction outut Histogram.(c) HE outut lot.(d) Contrast Stretching outut Histogram.(e) Virtual Enhanced outut Histogram. TABLE I (c) (e) Comare the MSE, PSNR and SSIM value. Methods Fig Name MSE Gamma Correction Histogram Equalization Contrast Stretching Virtual Enhanced PSNR SSIM Library image 29.940 23.402 0.509 Library image 76.358 18.802 0.357 Library image 67.139 19.895 0.575 Library image 142.06 15.587 0.159 IV. CONCLUSION [1] Hany Farid Blind Inverse Gamma Correction IEEE Transactions on Image Processing, May 2001. www.oynton.com. [2] Cristian Munteanu and Agostinho Rosa, Gray-Scale Image Enhancement as an Automatic Process Driven by Evolution, IEEE Transactions on Systems, vol. 34, no. 2, Aril 2004. [3] S.-D. Chen and R. Ramli "Contrast Enhancement Using Recursive Mean-Searate Histogram Equalization for Scalable Brightness Preservation," IEEE Transaction on Consumer Electronics, vol. 49, no. 4, 2003. [4] Gyu-Hee Par, Hwa-Hyun Cho, and Myung-Ryul Choi, A Contrast Enhancement Method using Dynamic Range Searate Histogram Equalization IEEE Transaction on Consumer Electronics, October 2008. [5] Soong-Der Chen, Abd. Rahman Ramli, Minimum Mean Brightness Error Bi-Histogram Equalization in Contrast Enhancement, IEEE Transactions on consumer Electronics, vol. 49, no. 4, November 2003. [6] J.Alex Star, Adative Image Contrast Enhancement Using Generalizations of Histogram Equalization IEEE Transactions on image rocessing, vol 9, no 5, May 2000. [7] Taeyung Kim and Jooni Pai, Adative Contrast Enhancement Using Gain-Controllable Clied Histogram Equalization IEEE Transactions on Consumer Electronics, vol 54, no 4, November 2008. [8] William K. Pratt, Digital Image Processing, John Wiley & sons, 2008. [9] S.Srinivasan and N.Balaram, Adative Contrast Enhancement Using Local Region Stretching Conference Paer. October 2006. [10] Milan Sona, Vaclav Hlavac and Roger Boyle, Image rocessing, Analysis and Machine vision, Broos/ Cole,2001. [11] Wilhelm Burger and Mar J Burger, Digital Image rocessing, Sringer, 2008. Raguathi S Obtained D.E.C.E in The Kavery Polytechnic College, year of assing 2007, B.E degree in Electronics And Communication Engineering from The Kavery Engineering College, 2010 and M.E in VLSI Design from K.S.Rangasamy College of Technology Anna University, Chennai, Tamil Nadu, India in 2013. His interesting Research area image enhancement and image segmentation. Santhi K comleted her D.E.E.E., B.E.,and M.E., in the year of 1993, 2000 and 2007 resectively.now she is a research scholar in Anna University,Chennai,India. Her research areas includes image enhancement, image segmentation and control systems. Currently she is woring as a Assistant Professor at the Electronics & Communication Engineering at K.S.Rangasamy College of Technology (Autonomous), Namaal,India. A new hybrid image enhancement aroach Parameter-Controlled Virtual Histogram Distribution (PCVHD) is driven to rocesses on luminance and chrominance comonents of the image simultaneously the overall contrast and the sharness of an image were enhanced. The roosed aroach introduces the arameters to increase the visibility of secified features ortion or asects of the image. Automatic rocess and also it has otential for various alications to enhance secific categories of images, such as surveillance videos/images, biomedical images and satellite images. 562