Research Paper A LOW CONTRAST COLOR IMAGE ENHANCEMENT BASED ON COLOR SPACE CONVERSION 2D-DWT AND SVD G. Saravanan* a and G.

Similar documents
A Modified SVD-DCT Method for Enhancement of Low Contrast Satellite Images

Satellite Image Processing Using Singular Value Decomposition and Discrete Wavelet Transform

An Optimal Gamma Correction Based Image Contrast Enhancement Using DWT-SVD

Image Contrast Enhancement in Wavelet Domain

ANALYSIS OF IMAGE ENHANCEMENT USING SVD DCT AND SVD DWT

Edge Preserving Contrast Enhancement Method Using PCA in Satellite Images

A Color Image Enhancement based on Discrete Wavelet Transform

IMAGE DIGITIZATION BY WAVELET COEFFICIENT WITH HISTOGRAM SHAPING AND SPECIFICATION

Image Resolution Improvement By Using DWT & SWT Transform

Document Text Extraction from Document Images Using Haar Discrete Wavelet Transform

HYBRID TRANSFORMATION TECHNIQUE FOR IMAGE COMPRESSION

CONTRAST ENHANCEMENT OF COLOR IMAGES BASED ON WAVELET TRANSFORM AND HUMAN VISUAL SYSTEM

Feature Based Watermarking Algorithm by Adopting Arnold Transform

A NEW ROBUST IMAGE WATERMARKING SCHEME BASED ON DWT WITH SVD

Enhancement of Low Contrast Satellite Images using Discrete Cosine Transform and Singular Value Decomposition

Digital Image Watermarking using Fuzzy Logic approach based on DWT and SVD

Robust Image Watermarking based on Discrete Wavelet Transform, Discrete Cosine Transform & Singular Value Decomposition

Palmprint Recognition Using Transform Domain and Spatial Domain Techniques

Curvelet Transform with Adaptive Tiling

Image Fusion Using Double Density Discrete Wavelet Transform

IMAGE COMPRESSION USING HYBRID TRANSFORM TECHNIQUE

Comparison of Wavelet Based Watermarking Techniques for Various Attacks

Image Contrast Enhancement by Scaling Reconstructed Approximation Coefficients using SVD Combined Masking Technique

IMPROVED MOTION-BASED LOCALIZED SUPER RESOLUTION TECHNIQUE USING DISCRETE WAVELET TRANSFORM FOR LOW RESOLUTION VIDEO ENHANCEMENT

Digital Image Steganography Techniques: Case Study. Karnataka, India.

An Effective Multi-Focus Medical Image Fusion Using Dual Tree Compactly Supported Shear-let Transform Based on Local Energy Means

Efficient Watermarking Technique using DWT, SVD, Rail Fence on Digital Images

IMAGE FUSION PARAMETER ESTIMATION AND COMPARISON BETWEEN SVD AND DWT TECHNIQUE

Biometric Security System Using Palm print

WAVELET SHRINKAGE ADAPTIVE HISTOGRAM EQUALIZATION FOR MEDICAL IMAGES

Robust color segmentation algorithms in illumination variation conditions

CHAPTER 3 WAVELET DECOMPOSITION USING HAAR WAVELET

Texture Analysis of Painted Strokes 1) Martin Lettner, Paul Kammerer, Robert Sablatnig

Robust Lossless Image Watermarking in Integer Wavelet Domain using SVD

Invisible Digital Watermarking using Discrete Wavelet Transformation and Singular Value Decomposition

Contrast Improvement on Various Gray Scale Images Together With Gaussian Filter and Histogram Equalization

A Robust Color Image Watermarking Using Maximum Wavelet-Tree Difference Scheme

Denoising and Edge Detection Using Sobelmethod

A DWT, DCT AND SVD BASED WATERMARKING TECHNIQUE TO PROTECT THE IMAGE PIRACY

INTERNATIONAL JOURNAL OF PURE AND APPLIED RESEARCH IN ENGINEERING AND TECHNOLOGY

DWT-SVD based Multiple Watermarking Techniques

PRINCIPAL COMPONENT ANALYSIS IMAGE DENOISING USING LOCAL PIXEL GROUPING

Study of Brightness Preservation Histogram Equalization Techniques

Image Compression & Decompression using DWT & IDWT Algorithm in Verilog HDL

Adaptive Wavelet Image Denoising Based on the Entropy of Homogenus Regions

Key Frame Extraction using Faber-Schauder Wavelet

International Journal of Research in Advent Technology Available Online at:

RKUniversity, India. Key Words Digital image processing, Image enhancement, FPGA, Hardware design languages, Verilog.

Image Watermarking with Biorthogonal and Coiflet Wavelets at Different Levels

Image Quality Assessment Techniques: An Overview

DWT-SVD Based Hybrid Approach for Digital Watermarking Using Fusion Method

FEATURE EXTRACTION TECHNIQUES FOR IMAGE RETRIEVAL USING HAAR AND GLCM

DIGITAL IMAGE HIDING ALGORITHM FOR SECRET COMMUNICATION

Image Enhancement Techniques for Fingerprint Identification

Generation of Digital Watermarked Anaglyph 3D Image Using DWT

Using Shift Number Coding with Wavelet Transform for Image Compression

Change Detection in Remotely Sensed Images Based on Image Fusion and Fuzzy Clustering

Resolution Magnification Technique for Satellite Images Using DT- CWT and NLM

RESTORATION OF DEGRADED DOCUMENTS USING IMAGE BINARIZATION TECHNIQUE

Dual Tree Complex Wavelet Transform (DTCWT) based Adaptive Interpolation Technique for Enhancement of Image Resolution

Based on Regression Diagnostics

Learning based face hallucination techniques: A survey

Handwritten Script Recognition at Block Level

Footprint Recognition using Modified Sequential Haar Energy Transform (MSHET)

Periodicity Extraction using Superposition of Distance Matching Function and One-dimensional Haar Wavelet Transform

FINGERPRINTING SCHEME FOR FILE SHARING IN TRANSFORM DOMAIN

Comparative Analysis of Different Spatial and Transform Domain based Image Watermarking Techniques

Image Enhancement Using Fuzzy Logic

SCALED WAVELET TRANSFORM VIDEO WATERMARKING METHOD USING HYBRID TECHNIQUE: SWT-SVD-DCT

A Nonoblivious Image Watermarking System Based on Singular Value Decomposition and Texture Segmentation

Intensification Of Dark Mode Images Using FFT And Bilog Transformation

DENOISING OF COMPUTER TOMOGRAPHY IMAGES USING CURVELET TRANSFORM

Constrained PDF based histogram equalization for image constrast enhancement

Image Classification Using Wavelet Coefficients in Low-pass Bands

Performance Evaluation of Fusion of Infrared and Visible Images

A Robust Digital Watermarking Scheme using BTC-PF in Wavelet Domain

International Journal of Advanced Research in Computer Science and Software Engineering

Digital Color Image Watermarking In RGB Planes Using DWT-DCT-SVD Coefficients

International Journal of Computer Science and Mobile Computing

Motivation. Intensity Levels

IMAGE DE-NOISING IN WAVELET DOMAIN

Robust Watermarking Method for Color Images Using DCT Coefficients of Watermark

Robust Image Watermarking based on DCT-DWT- SVD Method

An Approach for Reduction of Rain Streaks from a Single Image

COMPARATIVE STUDY OF IMAGE FUSION TECHNIQUES IN SPATIAL AND TRANSFORM DOMAIN

Patch-Based Color Image Denoising using efficient Pixel-Wise Weighting Techniques

Adaptive Quantization for Video Compression in Frequency Domain

Non-Linear Masking based Contrast Enhancement via Illumination Estimation

Motivation. Gray Levels

Image Compression Algorithm for Different Wavelet Codes

Sparse Representation Based Super-Resolution Algorithm using Wavelet Domain Interpolation and Nonlocal Means

Wavelet Based Image Retrieval Method

INTERNATIONAL JOURNAL OF PURE AND APPLIED RESEARCH IN ENGINEERING AND TECHNOLOGY

An ICA based Approach for Complex Color Scene Text Binarization

A Novel NSCT Based Medical Image Fusion Technique

A Novel Approach of Watershed Segmentation of Noisy Image Using Adaptive Wavelet Threshold

A New Approach to Compressed Image Steganography Using Wavelet Transform

Illumination-Robust Face Recognition based on Gabor Feature Face Intrinsic Identity PCA Model

IMAGE PROCESSING USING DISCRETE WAVELET TRANSFORM

International Journal of Advanced Research in Computer Science and Software Engineering

Transcription:

Research Paper A LOW CONTRAST COLOR IMAGE ENHANCEMENT BASED ON COLOR SPACE CONVERSION 2D-DWT AND SVD G. Saravanan* a and G. Yamuna b Address for Correspondence a Asst. Professor, Dept. of Electrical Engineering, Annamalai University, Annamalai Nagar, Tamil Nadu, India, 608002. b Professor, Dept. of Electronics and Communication Engineering, Annamalai University, Annamalai Nagar, Tamil Nadu, India, 608002. ABSTRACT: Digital Imaging systems are traditionally not good in low light conditions. In order to solve this issue, a new algorithm is proposed for contrast perfection. This technique first convert RGB color image to HSV, then the luminance part V decomposes into the four frequency sub- bands by using 2D-DWT. These techniques also apply the general histogram equalization (GHE) for V component then apply 2D-DWT it decomposes into four sub bands. To compute brightness improved LL sub band, normalize singular value matrix (SVD) obtained by both LL sub bands. The new LL sub band merged with high frequency sub bands, it reconstructs the image by applying 2D-IDWT. The new enhanced V combined with adaptive histogram equalized saturation part of S and H is unaltered due to color distortion and back to enhanced high contrast color image. The experimental results indicated that the image contrast improved by the proposed scheme is higher than the other methods and state-of-the-art techniques. KEYWORDS: Color Space Conversion (CSC), Discrete Wavelet Transform (DWT), Singular Value Decomposition (SVD), Histogram equalization, Low Contrast Enhancement. I INTRODUCTION Image enrichment is the development of improving appearance of the image for a specific application, widely used in the field of image processing [1]. Contrast enhancement is frequently referred to as one of the vital issues in image processing. Contrast is formed by the variation in luminance reflected from two adjoining surfaces. In visual observation contrast is find by the variation in the color and brightness of an object with other objects. If the contrast of an image is greatly intense on a specific range, the information may be lost, to solve this to optimize the contrast of an image so as to signify all the information in the input image. In many image processing applications to solve this issue, the HSV color space [2] is based on cylindrical coordinate is used. The HSV representation defines a color space in terms of three basic mechanisms. A new algorithm was stated that adaptive color image enrichment based on human visual properties in HSV space [3]. The Global Histogram Equalization (GHE) technique is one of the simplest and most effective primitives for contrast enhancement [4], which attempts to produce an output histogram that is uniform [5]. Gamma correction and Local Histogram Equalization (LHE) [6], Adaptive Histogram Equalization (AHE) computes several histograms, each corresponding to a distinct section of the image for contrast enhancement [7]. These techniques are very simple and effective Indies for the contrast enhancement. Brightness preserving histogram equalization for image contrast enhancement is the improvement in histogram equalization [8]. For further improvement of contrast enhancement method which is based on the Singular Value Decomposition of the Low-Low frequency sub band of the Discrete Wavelet Transform (DWT) [9]. Color image enhancement by Color Space Conversion HSV color space enhanced and inverted back to RGB color space [10, 11]. Currently, wavelets have been used rather frequently in image processing. It has been used for feature extraction [12], denoising [13], image equalization enhancement [14], and face recognition [15]. Previous literature focused on low contrast image enhancement in gray level images or the color image converted to grayscale image and then processed retrieved to color image. In order to solve this issue, a new method proposed for a low contrast image enhancement scheme with color Space Conversion (CSC). Fig. 1 Flow Diagram The rest of the paper is structured as follows. Section II discusses about the proposed color image enhancement scheme. Section III discusses the Experimental results obtained and Discussion and Section IV concludes the paper. II. METHODOLOGY A low contrast and dark color images, which has complete information but is not visible. Similarly satellite CT scans and satellite images are also dark images. So the enhancement of such images will help to get more details. The difficulty is how the contrast of an image can be enhanced from the input low contrast images and CT images. Here a new low contrast enhancement technique is proposed. There are 3 steps involved. First one is Color space conversion RGB to HSV so that the operation can be

done on luminance V part. Then DWT and followed by SVD then invert into back. The result shows that images are visibly enhanced using CSC-DWT-SVD method by incorporating AHE. As it was mentioned in the preface, the illumination information is embedded in LL sub band. The edges are concentrated in additional sub bands (LH, HL, and HH). Thus, extrication the high frequency sub bands and applying the illumination enhancement in LL sub band only, will protect the edge information from possible degradation. After reconstructing the final image by with IDWT, the consequential image will not only be improved with respect to illumination, but also it resolve be sharper. The common practice of the proposed scheme is as follows. The input image, A, is first processed by using GHE to generate. Then both of these images are transformed by DWT into four sub band images. The correction coefficient for singular value matrix is calculated by with the subsequent equation (1). Max( LL ) Max( LLA ) (1) Where is the LL singular value matrix of the LLA input image and LL is the LL band singular value matrix of the output of the GHE. The new LL image is composed by equation (2). LLA LLA (2) New LL A U LL V A LLA LL A Now the LL A and LH A, HL A, and HH A sub band images of the original image are recombined by applying IDWT, to generate the resultant equalized image A by equation (3). A IDWT (NewLL A, LH A, HL A, HH A ) (3) In this proposed work uses Haar wavelet function as the mother function of the DWT. In the subsequent section the investigational results and the comparison of the aforementioned usual and state-of-art techniques are discussed. 2.1 Algorithm The steps is as follows Step 1: Read the Low Contrast Color Image. Step 2: Color Space Conversion of RGB HSV. Step 3: Take the Luminance component V A and apply DWT of the image and decompose into LL A LH A HL A and HH A. Step 4: Apply the AHE to the image V to get enhanced V. Step 5: Take the DWT of the V image and decompose into LH and LL HL HH. Step 6: Take the SVD of both LL components. Step 7: Take the correction coefficient for the singular value. Step 8: Obtain the New LL image by multiplying U, correction coefficient, Σ,V values from SVD. Step 9: Apply the IDWT to New LL image and LH A HL A HH A and get enhanced V. Step 10: Apply the AHE for saturation part of S and H is unaltered. Step 11: Color Space Conversion HSV RGB, Get Enhanced color image. The above steps illustrated in Fig. 1. 2.2 Color Space Conversion 2.2.1. RGB HSV Color Space Conversion These primary colors can be collective to produce vast amount of secondary colors. Even though it is feasible to improve a digital actual color image by applying offered grey-level image improvement algorithms to every red, green and blue channel, the resultant image may not be improved optimally. RGB color space has limitation in representing shading property or fast illumination varying. So as to solve this trouble, converting an image from RGB space to HSV space. The HSV model defines a color space in terms of three basic mechanisms. Hue represents group of color such as red, blue, or yellow that falls from 0 to 360 degrees. Saturation is the vitality of color that falls from 0 to 100%. The lesser the saturation of a color, the extra gray the image looks and extra dull the color appears. Value is the brightness of the color that ranges from 0 to 100%. HSV space is as well recognized as hex cone color model. The HSV color space is extensively used to produce high quality digital graphics. The RGB to HSV conversion formula is shown by the equation (4). R R 255 Cmax C G ; G 255 MAX(R,G, B ; B 255 B ) min MIN(R,G,B ) Cmax C min G B 60 mod 6, Cmax R B R H 60 2, Cmax G R G 60 4, C max B (4) 0, 0 S, 0 Cmax V C max 2.2.2. Reverse Transformation: HSV to RGB Color Space Conversion The Reverse transformation by equation (5). C V X S hsv X C 1 (H / 60 ) mod 2 1 (0,0,0,) If H is Undefine (C,X,0) If 0 H60 (X,C,0) If 60 H120 ( R 1 G 1 B 1 ) (0,C,X) If 120 H 180 (0,X,C) If 180 H 240 (X,0,C) If 240 H300 (C,0,X) If 300 H360 (5) m V C ( R,G,B) (R 1 m, G 1 m, B 1 m) 2.3 Histogram Manipulation The histogram of an image is a plot of number of occurrences of gray levels in the image aligned with the gray level values. The histogram gives a suitable summary of the intensities in an image. Equalization is the procedure that attempts to stretch gray level in an image so that they are evenly distributed across

their range. Histogram equalization reallocates the brightness values of pixels based on image enhancement. Histogram equalization gives more visually satisfying results. Adaptive Histogram Equalization (AHE) calculates numerous histograms, each equivalent to a different segment of the image, and uses them to reallocate the lightness values of the image. It is therefore suitable for make better the neighboring contrast of an image. Adaptive histogram equalization transforms every pixel with a alteration function resulting from a neighborhood region. It operates on small data regions (tiles) and each tile's contrast is improved, in order that the histogram of the output area roughly matches the specified histogram. The neighboring tiles are then collective with bilinear interpolation that eliminates artificially induced boundaries. The contrast, especially in uniform areas, can be restricted in order to evade amplifying the noise which might be there in the image. 2.4 2-D Discrete Wavelet Transforms The second stage of this scheme uses 2-D Discrete Wavelet Transformation (DWT) at first level decomposition. DWT converts the image from the spatial field to frequency field. The image can be alienated with four elements those are LL, LH, HL and HH as shown in Fig. 2. In additional, those four portions are represented four frequency regions in the image. For the low-frequency domain LL is sensitively with human eyes. the image and the standard deviation represent (σ) the contrast present in the images. The proposed enhancement scheme represents the better contrast as well as better brightness with suitable contrast. Though, the approximated mean (μ) and standard deviation (σ) in Fig. 3 of the proposed scheme covers a good range of the better illumination. Therefore the observation of the proposed scheme gives the better consequence. In order to demonstrate the advantage of the proposed methodology four different images have been taken for analysis. The singular values designate luminance of every image layer after decomposition using 2D-DWT methodology for LL Sub band because it has the image approximation. The Mean (μ) & standard deviation ( ), MSE and PSNR values are given below for analysis of this result. The quality measure of the images calculated using the table 1 the proposed scheme provides better contrast as well as better brightness as shown in table 2. Fig. 4 illustrates the output of proposed method with other methods. Table 1 Quality Measures Fig. 2 Frequency distribution of DWT 2.5 Single value Decomposition (SVD) SVD is based on a theorem as of linear algebra which states that a rectangular matrix A, which is a multiplication of three matrices that is (i) an orthogonal matrix UA, (ii) a diagonal matrix ΣA and (iii) the invert of an orthogonal matrix VA. The singular-value-based image equalization (SVE) technique is modeled on equalizing the singular value matrix gained by singular value decomposition (SVD). The singular value matrix represents the intensity information of input image and some change on the singular values alters the intensity of the input image. SVD of an image, which can be interpreted as a matrix is written using the equation (6) as follows A V T UA A A (6) Where U A and V A are orthogonal square matrices identified as hanger and aligner correspondingly, and Σ A matrix holds the arranged singular values on its main diagonal. The scheme of using SVD for image equalization obtained from this information that Σ A contains the intensity information of the given image. 2.6 Performance Measures The performance of this method is measured in terms of following significant parameters, the superiority of the visual consequences designates that the proposed scheme is brighter and more contrast than existing schemes as compared. After getting mean and standard deviation, it is found that the proposed algorithm gives better results in comparison with the existing schemes. Mean (μ) represent the intensity of Fig. 3 Output of Low contrast image Enhanced by various Histogram Equalization methods and Proposed Scheme

and a small amount improved in brightness (Mean) and contrast (standard deviation). Even though the LHE gives good PSNR but comparing with other quality measures proposed scheme has good enhanced output. Table 3 shows the comparison of quality measure for proposed CSC-DWT-SVD low-contrast enhancement scheme with other methods and Fig. 5 and 6 shows the graphical representation of comparison of EME and AMBE values respectively. Table 3 Comparison of Quality Measure for Proposed CSC-DWT-SVD scheme with other Methods Fig. 4 (a) Input image lady (b) Input image doll (c) Input image couple (d)-(f) Enhancement results of Naik and Murthy s method (g)-(i) Enhancement results of Ancuti and Ancuti s method (j)-(l) Enhancement results of the proposed CSC,DWT- SVD method III EXPERIMENTAL RESULTS AND DISCUSSIONS The quality of the visual results indicates that the proposed equalization technique is and brighter and more contrast than the one achieved by GHE, LHE, and AHE. The resultant image generated by the above schemes is differentiating with the image attained by the proposed method. In order to show the superiority of the proposed method over the GHE LHE and AHE Fig. 3 have been generated, it illustrates low contrast images have been balanced by using GHE, LHE, AHE and the suggested enhancement scheme. Table 2 Quality Measures for various Histogram Equalization methods and Proposed Scheme Naik and Murthy [17] developed scheme is used to generalize the histogram equalization method for grey scale images to color images and bypass the CSCs for image enhancement. This method generalizes the grey scale contrast intensification techniques to color images and produces the lowvalue for measure of enhancement. Ancuti and Ancuti [18] developed method is a fusion-based approach that obtains from two original hazy image inputs by employing a white balance and a contrast enhancing scheme, it requires two original degraded images to enhance the system. The proposed CSC- DWT- SVD Low-contrast enhancement scheme gives high-value for measure of enhancement (EME) and maintains lesser value of Absolute Mean Brightness Error (AMBE) on comparing with the existing methods. Fig. 5 Comparison of EME Values for Proposed CSC- DWT-SVD scheme by other Methods The quality measure results are shown in Table 2 for various Histogram Equalization methods and Proposed scheme for color images. From the result, the proposed scheme has less MSE and Improved PSNR, Mean and Standard deviation when compared with other techniques and regarding PSNR, Local Histogram Equalization (LHE) scheme has higher Fig. 6 Comparison of AMBE Values for Proposed CSC- DWT-SVD scheme by other Methods IV. CONCLUSION In this work, a new image enhancement scheme based on CSC, DWT and SVD was proposed. The proposed technique converted the image from RGB color space to HSV color space and luminance part V in spatial domain into the DWT domain and after equalizing the singular value matrix of the LL sub

band image, it restructured the image in the spatial domain by using IDWT. The Saturation element S was improved by AHE. The scheme was evaluated with the GHE, LHE and AHE techniques. The experimental results were illustrating the superiority of the proposed scheme over the state-of-art techniques and other methods. This work can be extended by converting these files into VHSIC hardware description language (VHDL) and implementing through field programmable gate array (FPGA) processor. REFERENCES 1. R. C. Gonzalez, and R. E. Woods, Digital Image Processing, Englewood Cliffs, NJ: Prentice-Hall, 2007. 2. Dong Yu, Li-Hong Ma and Han-Qing Lu, Normalized SI Correction for Hue-Preserving Color Image Enhancement, Proceedings of the 6th International Conference on Machine Learning and Cybernetics (ICMLC 2007), Vol. 3, pp. 14981503, 2007. 3. Gang Song and Xiang-Lei Qiao, Adaptive Color Image Enhancement based on Human Visual Properties, Proceedings of International Conference on Image and Signal Processing, 2008. 4. T. Kim and H. S. Yang, A multidimensional histogram equalization by fitting an isotropic Gaussian mixture to a uniform distribution, in Proc. IEEE Int. Conf. Image Process., Oct. 8 11, pp. 2865 2868, 2006. 5. A. R. Weeks, L. J. Sartor, and H. R. Myler, Histogram specification of 24-bit color images in the color difference (C-Y) color space, Proc. SPIE, vol. 3646, pp. 319 329, 1999. 6. S. Chitwong, T. Boonmee, and F. Cheevasuvit, Enhancement of color image obtained from PCA-FCM technique using local area histogram equalization, Proc. SPIE, vol. 4787, pp. 98 106, 2002. 7. T. K. Kim, J. K. Paik, and B. S. Kang, Contrast enhancement system using spatially adaptive histogram equalization with temporal filtering, IEEE Trans. Consum. Electron., vol. 44, no. 1, pp. 82 87, Feb. 1998. 8. H. Ibrahim and N. S. P. Kong, Brightness preserving dynamic histogram equalization for image contrast enhancement, IEEE Trans. Consum. Electron., vol. 53, no. 4, pp. 1752 1758, Nov. 2007. 9. H. Demirel, G. Anbarjafari, and M. N. S. Jahromi, Image Equalization Based On Singular Value Decomposition, Proceeding of IEEE Conference on Computer and Information Sciences, pp. 1-5, 2008. 10. G.saravanan, G Yamuna and R.vivek. Article: A Color Image Enhancement based on Discrete Wavelet Transform. IJCA Proceedings on National Conference on Emerging Trends in Information and Communication Technology 2013 NCETICT:1-8, 2014. 11. Marc Strunkmann and Ulf Witkowski Tim Kaulmann, "FPGA-Based Object Detection in Robot Soccer Application," in Proceedings of the 3rd International Symposium on Autonomous Mini robots for Research and Edutainment, pp. 135-140, 2006. 12. J. L. Starck, E. J. Candes, and D. L. Donoho, The curvelet transform for image denoising, IEEE Transactions on Image Processing, Vol. 11, pp: 670-684, 2002. 13. J. W. Wang and W. Y. Chen, Eye detection based on head contour geometry and wavelet sub band projection, Optical Engineering, Vol. 45, No. 5, 2006. 14. H. Demirel, C. Ozcinar, and G. Anbarjafari, Satellite Image Contrast Enhancement Using Discrete Wavelet Transform and Singular Value Decomposition, IEEE Geosciences and Remote Sensing Letters, Vol. 7, No. 2, pp. 334-338, 2010. 15. C. C. Liu, D. Q. Dai, and H. Yan, Local discriminant wavelet packet coordinates for face recognition, Journal of Machine Learning Research, Vol. 8, pp: 1165-1195, 2007. 16. S. Agaian, B. Silver and K. Panetta, Transform Coefficient Histogram-Based Image Enhancement Algorithms using Contrast Entropy, IEEE Transactions on Image Processing, Vol. 16, No. 3, pp. 741-757, 2007. 17. S.J. Naik, and C.A. Murthy, Hue-Preserving Color Image Enhancement without Gamut Problem, IEEE Transactions on Image Processing, Vol. 12, No. 12, December, pp. 1591 1598, 2003. 18. C.O. Ancuti, and C. Ancuti, Single Image Dehazing by Multi-Scale Fusion, IEEE Transactions on Image Processing, Vol. 22, No. 8, August, pp. 3271-3282, 2013.