Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 1
ISS N 2229-5518
Brightness Preserving Image Enhancement Using Modified Dualistic Sub Image Histogram Equalization
Mrs. Ashwini Sachin Zadbuke
image. Th is paper provides the modif ied dualistic sub image HE method w hich preserves the brightness of the image.
—————————— ——————————
The goal of image enhancement techniques is to improve a quality of an image such that enhanced image is better than the original image. Several image enhancement techniques have been proposed in both spatial and transform domains. In the spatial domain techniques, intensity values of images have been modified whereas in the transform domain techniques, transform domain coefficients are modified, typically, scaled.
Histogram equalization (HE) [1] is a simple and effective contrast enhancement technique which distributes pixel values uniformly such that enhanced image have linear cumulative histogram. The HE technique is a global operation hence; it does not preserve the image brightness. To overcome this issue local-HE [2] and brightness preserving local HE [3]-[13] techniques have been proposed. In Brightness preserving bi histogram equalization (BBHE) [3] and dualistic sub-image histogram equalization (DSIHE) [4] techniques,
A histogram has been divided into two sub-histograms such that one contains high intensity pixels and another contains low intensity pixels. And then equalize each part independently with HE technique. The BBSE and DSIHE techniques use mean and median values, respectively as separation intensity to divide the histogram into two sub- histograms. Minimum mean brightness error bi-histogram equalization (MMBEBHE) [5] is an extension of the BBHE technique.
Hence it is inappropriate to explicitly conclude which method actually demonstrates the best performance.
So the scope of this paper is to highlight only the methods used for brightness preserving image enhancement using histogram equalization.
This paper is organized as follows: section II explains
various algorithms employed for brightness preserving
image enhancement. Section III explains different criterion
for brightness preservation as well as image enhancement, section V explains the test images used Section VI gives results of method in tables section VII
Shows resultant output images section VIII concludes the paper followed by references.
[1] HE (Histogram equalization):
In this section, we briefly describe conventional histogram equalization and its variant methods. In what follows, we will use the symbols and notations similar to the ones in.
• X = {X(i, j)} is an image with L discrete gray levels {X0,
X1, ⋅⋅⋅, XL-1}, where X (i, j) is the intensity of the image at the 2D position (i, j) and X(i, j) ∈ {X0, X1, ⋅⋅⋅, XL-1}.
• H(X) = {n0, n1, ⋅⋅⋅, nk, ⋅⋅⋅, nL-1} is the image X’s histogram,
where nk is the number of pixels whose gray level is Xk.
• XM: the mean brightness of the image X
• XG: the middle gray level of the image X, i.e., (X0+XL-
1)/2.
Conventional Histogram Equalization:
IJSER © 2012
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 2
ISS N 2229-5518
Consider the input image X. Based on the histogram H(X), the probability density function (PDF) of the image is defined
P (k) =k/N=k/ n0+n1+n2+n3+…. +nL-1 for k=0, 1, 2…
L-1; (1)
Where N is the total number of pixels in the image. From the PDF in (1), the cumulative distribution function (CDF) is defined as
C (k) = for k=0, 1… L-1. (2) Note that c (L-1) = 1 from (1) and (2). Based on the CDF,
histogram equalization now maps an input gray level Xk
into an output gray level f (k), where f (k), commonly called
a level transformation function, is defined as
(k) =X0+ (XL-1-X0).c (k). (3) Thus, histogram equalization remaps the input image
into the entire dynamic range [X0, XL-1]. Note also that
theoretically conventional histogram equalization produces
the output image whose mean brightness is XG regardless of the brightness of the input image.
[II] BBHE (Brightness Preserving Bi Histogram Equalization):
BBHE first decomposes the input histogram H(X) into two Sub-histograms HL(X) and HU(X) by using the input mean XM, Where HL(X) is associated with the gray levels
{X0, X1, ⋅⋅⋅, XM}: and HU(X) is associated with the gray levels {XM+1 , XM+2 , ⋅⋅⋅, XL-1}. Then it performs
conventional histogram equalization on HL(X) and HU(X)
independently. It is shown that if the histogram H(X) has a
symmetrical distribution around XM, the mean brightness
of the output image is (XM + XG)/2.
[III] BPDHE: (Brightness Preserving Dynamic Histogram
Equalization):
This is proposed in this paper consists of five steps:
1. Smooth the histogram with Gaussian filter.
2. Detection of the location of local maximums from the smoothed histogram.
3. Map each partition into a new dynamic range.
4. Equalize each partition independently.
5. Normalize the image brightness.
[IV] MDSIHE (MODIFIED Dualistic Sub-Image Histogram
Equalization):
If a gray level XD satisfies c (XD) = 0.5, then it is called the median of the image X. DSIHE is similar to BBHE except that the threshold for histogram segmentation is the median XD of
the input image. That is, the input histogram H(X) is partitioned into two sub-histograms HL(X) and HU(X) not
by the input mean XM, but by the input median XD. Each of HL(X) And HU(X) is then equalized independently as BBHE.Before histogram equalization histogram normalization is done.
To evaluate the effectiveness we chose three widely-used metrics, i.e., AMBE (Absolute Mean Brightness Error), PSNR (Peak Signal-to-Noise Ratio), and entropy.
They are described in detail below.
1] ABSOLUTE MEAN BRIGHTNESS ERROR:
AMBE(X, Y) = |XM - YM|, where XM is the mean of the input image X = {X (i, j)} and YM is the mean of the output image Y = {Y (i, j)}.
2] INSPECTION OF VISUAL QUALITY:
In addition to the quantitative evaluation of brightness preservation using the AMBE values, it is also important to qualitatively assess. The major goal of the qualitative assessment is to judge if the output image is visually acceptable to human eyes and has a natural appearance.
1] PEAK SIGNAL TO NOISE RATIO:
Assuming that N is the total number of pixels in the input or output image, MSE (Mean Squared Error) is calculated through.
PSNR=10 log10(L-1)2/MSE
Based on MSE, PSNR is then defined as.
Note that the greater the PSNR, the better the output image quality.
MSE= ∑i∑j │X (i,j)-Y (i,j)2 │ ∕ N
2] ENTROPY:
For a given PDF p, entropy Ent[p] is computed. In general, the entropy is a useful tool to measure the Richness of the details in the output image.
Ent[p] = -∑k=0 (k)log2 p (k)
3] INSPECTION OF VISUAL QUALITY:
In addition to the quantitative eva luation of contrast enhancement using the PSNR and entropy values, it is also important to qualitatively assess the contrast enhancement.
IJSER © 2012
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 3
ISS N 2229-5518
The major goal of the qualitative assessment is to judge if the output image is visually acceptable to human eyes and has a natural appearance.
Image 1 57.95 58.29 33.73 11.84 7.184
Image 2 57.57 58.70 27.73 16.75 8.316
Image 3 31.26 33.79 34.68 7.097 7.873
( 1 ) (2)
Image 4 67.00 69.93 25.54 12.93 2.351
Image 5 64.91 65.73 23.46 17.64 1.848
Image 1 0.0213 0.0264 0.0016 0.9942 0.9998
(3 ) (4)
(5)
Image 2 0.0200 0.01777 0 0.9817 0.9995
Image 3 0.7202 0.7208 0 0.9998 0.9994
Image 4 0.0096 0.0097 0 0.9770 0.9999
Image 5 0.0187 0.0132 0 0.9453 0.9994
B] PSNR measurement:
IJSER © 2012
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 4
ISS N 2229-5518
C] Entropy measurement:
IMAGE | GLOBAL HE | LOCAL HE | ADP HE | BBHE | MDSIH E |
Image 1 | 11.99 | 11.383 | 15.29 | 15.826 | 16.02 |
Image 2 | 11.2717 | 9.9791 | 16.49 | 13.961 | 14.10 |
Image 3 | 16.0063 | 14.029 | 15.11 | 18.258 | 18.18 |
Image 4 | 11.1001 | 10.028 | 16.94 | 16.939 | 17.87 |
Image 5 | 10.7025 | 9.6213 | 17.67 | 16.393 | 16.80 |
AVG | 12.2141 | 11.008 | 16.304 | 16.275 | 19.975 |
* Input Image 1
IJSER © 2012
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 5
ISS N 2229-5518
Here we have discussed results of first five methods that are available for contrast enhancement & brightness preservation such as conventional global HE, local HE, ADPHE, BBHE, DSIHE.
The last method as MDSIHE gives better results than all other.
[1] A. Rafael C. Gonzalez, and Richard E. Woods, “Digital
Image Processing,” 2nd edition, Prentice Hall, 2002.
[2] Joung-Youn Kim, Lee-Sup Kim and Seung-Ho Hwang,
“An advanced contrast enhancement using partially
overlapped sub-block histogram equalization,” IEEE Trans.
Circuits Syst. Video Technol., vol. 11, no. 4, pp. 475 -484, Apr. 2001.
[3] Yeong-Taeg Kim, “Contrast enhancement using brightness preserving bi histogram equalization,” IEEE Trans. Consumer
Electronics, vol. 43, no. 1, pp. 1 -8, Feb. 1997.
[4] Yu Wan, Qian Chen and Bao-Min Zhang, “Image enhancement based on equal area dualistic sub-image histogram equalization method,” IEEE Trans. Consumer Electron., vol. 45, no. 1, pp. 68 -75, Feb. 1999.
[5] Soong-Der Chen and Abd. Rahman Ramli, “ Minimum mean brightness error bi-histogram equalization in contrast enhancement,” IEEE Trans. Consumer Electron., vol. 49,
no. 4, pp. 1310-1319, Nov. 2003.
[6] Soong-Der Chen and Abd. Rahman Ramli, “Contrast enhancement using recursive mean -separate histogram equalization for scalable
[7] Brightness preservation,” IEEE Trans. Consumer
Electron., vol. 49, no.4, pp. 1301 -1309, Nov. 2003.
[8] K. S. Sim, C. P. Tso and Y. Y. Tan, “Recursive sub-image histogram equalization applied to gray scale images,” Pattern Recognition Letters, vol. 28, no. 10, pp. 1209 -1221,
2007.
[9] D. Menotti, L. Najman, J. Facon and A. A. Araujo, “Multi-histogram equalization methods for contrast enhancement and brightness.
[10] H. Ibrahim and N. S. P. Kong, “Brightness preserving dynamic histogram equalization for image contrast enhancement,” IEEE Trans. Consumer Electron., vol. 53, no.
4, pp. 1752-1758, Nov. 2007.
[11]Nyamlkhagva Sengee and Heung Kook Choi, “Brightness Preserving Weight Clustering Histogram Equalization,” IEEE Trans. Consumer Electron., vol. 54, no.
3, pp. 1329-1337, Aug. 2008.
IJSER © 2012
Inte rnatio nal Jo urnal o f Sc ie ntific & Eng inee ring Re se arc h, Vo lume 3, Issue 2, February -2012 6
ISS N 2229-5518
[12]Hojat Yeganeh, Ali Ziaei and Amirhossein Rezaie, “A novel approach for contrast enhancement based on histogram equalization,”
In Proceedings of the International Conference on
Computer and Communication Engineering, pp. 256 -260,
2008.
[13]S.Aghagolzadeh and O. K. Ersoy, “Transform image enhancement,” Opt. Eng., vol. 31, pp. 614626, Mar. 1992.
[14] J. Tang, E. Peli and S. Acton, “Image enhancement using a contrast measure in the compressed domain,” IEEE Signal Process. Let., vol. 10, no. 10, pp. 289 -292, Oct. 2003.
[15] S. Lee, “An efficient content-based image enhancement in the
compressed domain using retinex theory,” IEEE Trans. Circuits Syst. Video Technol., vol. 17, no. 2, pp. 199 -213, Feb. 2007.
[16]Jayanta Mukherjee and Sanjit K. Mitra, “Enhancement of color images by scaling the DCT Coefficients,” IEEE Trans. Image
Process., vol. 17, no. 10, pp. 1783 -1794, Oct. 2008.
[17] X. Zong, A. F. Laine, E. A. Geiser and D. C. Wilson, “Denoising and contrast enhancement via wavelet shrinkage and nonlinear adaptive gain,” Proc. of SPIE, vol.
2762, pp. 566-574, Orlando, Apr. 1996.
[18] K. V. Velde, “Multi-scale color image enhancement,” in
Proc. Int. Conf Image Processing, vol. 3, pp. 584 -587, 1999. [19] Y. Wan and D. Shi, “Joint exact histogram specification and image enhancement through the wavelet transform,”
IEEE Trans. Image Process., vol. 16, no. 9, pp. 2245 -2250,
Sep.2007.
[20] J. Starck, F. Murtagh, E. J. Cands and D. L. Donoho, “Gray and color image contrast enhancement by the curvelet transform,” IEEE Trans. Image process., vol. 12, no.
6, pp. 706-717, Jun. 2003.
[21] Qingwu Li, Xue Ni and Guogao Liu, “Ceramic image processing using the second generation curvelet transform and watershed
algorithm,” In Proceedings of the International Conference
on Robotics and Biomimetics, pp. 2037 -2042, 2007.
[22] Ehsan Nezhadarya and Mohammad B. Shamsollahi, “Image contrast enhancement by contourlet transform,” In Proceedings of the 48th International Symposium, pp. 81 -
84, Jun. 2006.
[23] D. L. Donoho and M. R. Duncan, “Digital curvelet transform: Strategy, implementation and experiments,” Proc. SPIE, vol. 4056,
pp. 12-29, 2000.
[24] E. J. Cand`es, L. Demanet, D. L. Donoho and L. Ying, “Fast discrete curvelet transforms,” SIAM Multiscale Modeling Simul., 2006.
About Author:
Prof. Ashwini Sachin Zadbuke
M.E.(Digital Systems)
S.B.Patil College Of Engineering, Indapur. Dist.: Pune, (M.S) India.
IJSER © 2012