Article Preview
TopIntroduction
Image enhancement is the processing of an image using some transformation function such that the output image is more useful than the original one for some specific kind of applications (Gonzalez, Woods, 2002). The transformation function will change depending on the objective of image enhancement. Enhancement is a pre-processing step because different kinds of images suffer from poor contrast. Image enhancement can be applied for various image processing applications like contrast enhancement, noise reduction, edge enhancement and edge restoration (Gonzalez, Woods, 2002). The approaches for image enhancement are classified into two main categories: Spatial Domain methods where pixel values have been taken into consideration and Frequency Domain methods where Fourier transforms has been applied on the image (Gonzalez, Woods, 2002). Different spatial domain techniques are image negatives, log transformation, power-law transform, contrast stretching, gray-level slicing etc. (Gonzalez, Woods, 2002; Gorai, & Ghosh, 2009). Basically spatial domain image enhancement techniques divided into two categories: a) global enhancement methods b) local enhancement methods (Gonzalez, Woods, and 2002). Global enhancement methods like traditional histogram equalization (HE) only take global information of the image to enhance it. It does not use the local information to enhance the image. These methods take less time but over-enhancement is the main disadvantage (Gonzalez, Woods, 2002). Local enhancement methods like adaptive histogram equalization (AHE) produce good result than traditional HE but takes more time (Gonzalez, Woods, 2002; Gorai, & Ghosh, 2009). To overcome this problem a parameterized transfer function has been developed which consider global as well as local information to enhance the image (Gorai, & Ghosh, 2009; Munteanu, & Rosa, 2001). In this study this technique has been taken into consideration. Another most popular and important spatial domain image enhancement technique is histogram equalization which solely depends on the information of the histogram of the image (Gonzalez, Woods, 2002; Gorai, & Ghosh, 2011). Histogram equalization (HE) takes the cumulative histogram of the input image and dispenses the pixel values over the intensity range of the output image (Gonzalez, Woods, 2002; Cheng, & Shi, 2004; Chen & Ramli, 2004; Chen & Ramli, 2003; Chen & Ramli, 2003). Histogram equalization flattens the histogram of the image and contrast of the image has been increased due to this fact (Gonzalez & Woods, 2002). HE has been used in medical image processing, satellite image processing field. The main feature of HE is that the mean brightness of the enhanced image using HE tends to around the middle of the gray scale regardless of input image mean (Chen & Ramli, 2004). This fact is useful for face detection (Chen & Ramli, 2004) but does not suit for consumer electronics applications (Chen & Ramli, 2004). If the original brightness does not preserves then annoying artifacts and unnatural enhancement is happened to the output image (Chen & Ramli, 2004). To overcome this problem several modification has been done on traditional HE (Cheng & Shi, 2004; Chen & Ramli, 2004; Chen & Ramli, 2003; Chen & Ramli, 2003). Brightness Preserving Bi-Histogram Equalization (BBHE) is one of the simple variant of basic HE (Chen & Ramli, 2004; Chen & Ramli, 2003; Chen & Ramli, 2003). In BBHE histogram of the image separated around its mean and then the two divided parts equalized separately (Chen & Ramli, 2004; Shanmugavadivu, Balasubramanian, & Muruganandam, 2014). It is proved that BBHE method preserve the original brightness to a certain level (Chen & Ramli, 2004; Shanmugavadivu, Balasubramanian, & Muruganandam, 2014). There is another method known as Dualistic Sub-Image Histogram Equalization (DSIHE) Proposed by Wan (Chen & Ramli, 2004; Shanmugavadivu, Balasubramanian, & Muruganandam, 2014). The procedure of DSIHE is same as BBHE but the histogram is separated by median instead of mean (Shanmugavadivu, Balasubramanian, & Muruganandam, 2014; Chen & Ramli, 2004; Chen & Ramli, 2003; Chen & Ramli, 2003). DSHIE method is suitable for only those kinds of images which have uniform intensity distribution. The potential of preserving the original brightness is not so good of DSIHE method. There is another variant of BBHE method which is known as Minimum Mean Brightness Error Bi-Histogram Equalization (MMBEBHE) proposed by Chen and Ramli (Chen & Ramli, 2004; Chen & Ramli, 2003). In MMBEBHE, at first the histogram has been separated using a specified threshold which preserves the minimum mean brightness error between input and output image. Then the two parts have been equalized independently. This technique is better than BBHE but it also suffers from deficiency of contrast and brightness (Shanmugavadivu, Balasubramanian, & Muruganandam, 2014). Recursive Mean Separate Histogram Equalization (RMSHE) is another HE based procedure proposed by Chen and Ramli where the input image is separated recursively (Shanmugavadivu, Balasubramanian, & Muruganandam, 2014; Chen & Ramli, 2004; Chen & Ramli, 2003). Each part equalized independently and combines them to give an enhanced image. It is proved that this RMSHE method is better than BBHE (Chen & Ramli, 2004). RMSHE has been preserved the brightness well and gives a natural enhancement (Chen & Ramli, 2004). The disadvantages of RMSHE are large time complexity and finding of the optimal recursion level. So, all the above discussed HE variants suffer from original brightness preservation problem and unwanted artifacts. The main reason for these drawbacks is that these techniques do not change the original probability density function (PDF) or normalized histogram. As the PDF does not change the high probable intensity levels are over-enhanced. The simple weighted and thresholded technique has been developed in (Wang & Ward, 2007; Sengee & Choi, 2008; Kim & Chung, 2008) to change the PDF. In (Wang & Ward, 2007) a fast and effective image and video contrast enhancement technique has been proposed which is known as Weighted Threshold HE (WTHE). In this technique the PDF of the image is modified by weighting and thresholding prior to HE. A mean adjustment factor is also added to normalize the luminance changes. The advantages of that method are adaptivity of different images and control over the enhancement. These advantages are very difficult to achieve in traditional HE method and it’s above discussed variants like BBHE, DSIHE and MMBEBHE. There are also another two weighted methods are also developed in (Sengee & Choi, 2008; Kim & Chung, 2008). In (Sengee & Choi, 2008) a weighted clustering HE (WCHE) and in (Kim & Chung, 2008) a recursively separated and weighted HE (RSWHE) is proposed. These methods also have their own advantages. There are two parameters in WTHE (Wang & Ward, 2007). One is used to choose proper threshold and another is power law coefficient which is used to modify the PDF. The results are different depending on these parameters. If the parameters are chosen properly then it gives much better results which do not include over-enhancement or artifacts. So the efficiency of that method crucially depends on these parameters. The only demerit of WTHE is parameters have been set manually. To give the power of automation to this method or to choose the parameters optimally Artificial Bee Colony (ABC) algorithm has been used in this study.