(1)
The image enhancement techniques have become an important pre- processing tool for digital vision processing applications, where one of the most common degradations in images is their poor contrast quality. This suggests the use of contrast enhancement methods as an attempt to modify the intensity distribution of the image. The main objective of this paper is the analysis of mathematical morphological approach with comparison to various other state-of-art techniques for addressing the problems of low contrast in images. A technique used for contrast enhancement is the combined use of the top-hat bottom-hat transforms and logical operations. Histogram Equalization (HE) is one of the common methods used for improving contrast in digital images. This method is simple and effective for global contrast enhancement of images, but it suffers from some drawbacks. Contrast Limited Adaptive Histogram Equalization (CLAHE) enhances the local contrast of the images without the amplification of the noise. The proposed technique is best while compared qualitatively and quantitatively with existing technique.
Today, there is almost no area of technical endeavour that is not impacted in some way or another by digital image processing. The area of digital image processing is a dynamic field and new techniques and applications are reported routinely in professional literature and in new product announcements. Digital images are subject to a wide variety of distortions which may result in visual quality degradations. Image enhancement is crucial for many image processing applications. The ultimate goal of image enhancement techniques is to improve the visual information of a degraded image in a subjective process. Image sharpening is a classic problem in the field of image enhancement.
The principal objective of image sharpening is to highlight fine details in an image or to enhance details that have been blurred, either in error or as a natural effect of a particular method of image acquisition. Usages of image sharpening vary and include applications ranging from document and medical imaging to industrial inspection and autonomous guidance in military systems [1]. Linear operators have been the dominating filter class throughout the history of image processing. This is triggered by the computational efficiency of linear filtering algorithms. Despite the elegant linear system theory, not all image sharpening problems can be satisfactorily addressed through the use of linear filters. Many researchers now hold the view that it is not possible to obtain major breakthroughs in image sharpening without resorting to nonlinear methods[2]. Low contrast image results due to low light conditions, and lack of dynamic range of the camera sensor. Contrast stretching operation results in good quality image. For this, there have been several methods such as General Histogram Equalization (GHE) [3,7], Dynamic Histogram Equalization (DHE) [4], Dynamic Histogram Specification (DHS) [5], and Brightness preserved Dynamic Histogram Equalization [6]. GHE is a common technique for enhancing the appearance of the image. It involves finding a grey scale transformation function that creates an output image with a uniform histogram. DHE obtained from dynamic histogram specification generates the specified histogram dynamically from the input image.
The purpose of this paper is to give a comparative overview along with the discussion of visual results of several techniques used for contrast enhancement of images. To get the best result, maximize the contrast of the objects of interest to minimize the number of valleys found by the Watershed transform. The top-hat transform is defined as the difference between the original image and its opening. The opening of an image is the collection of foreground parts of an image that fit a particular structuring element. The bottom-hat transform is defined as the difference between the closing of the original image and the original image. The closing of an image is the collection of background parts of an image that fit a particular structuring element.
Section 1 provides objectives of Contrast Enhancement. Section 2 contains some of the existing techniques and their drawbacks. Methodology of present work has been described in section 3. Computer simulated results are in presented in Section 4, Section 5 provides a discussion of the visual results. Finally, the study is concluded.
Image enhancement is a pre-processing step in many image processing applications. The main objective of image enhancement is to improve the interpretability or perception of information in images for human viewers, or to provide better input for other automated image processing techniques. There are various reasons for poor quality of an image such as distortion being introduced by the imaging systems, lack of expertise of the operator or the adverse external conditions at the time of image acquisition. Mainly, Image enhancement includes intensity and contrast manipulation, noise reduction, edges sharpening and filtering, etc. Contrast Enhancement is focused on the problem of improving the contrast in an image to make various features more easily perceived. Contrast of an image is determined by its dynamic range, which is defined as the difference between lowest and highest intensity level. Contrast enhancement techniques have various application areas for enhancing the visual quality of low contrast images. Many contrast enhancement algorithms have been developed over the years. Contrast enhancement algorithms can broadly be divided into two categories: spatial domain techniques and frequency domain techniques.
In spatial domain techniques, the image enhancement is based on direct manipulation of the pixels in an image. Frequency domain processing techniques are based on modifying the Fourier transform of an image. In frequency domain methods, the image is first transferred in to frequency domain. It means that, the Fourier Transform of the image is computed first. All the enhancement operations are then performed on the Fourier transform of the image and then the Inverse Fourier transform is performed to get the resultant image. Contrast enhancement is one of the important research issues of image enhancement. An image taken in a very dark or bright situations result in low contrast image. However, the contrast of an image can be improved from the input image which has complete information but are not visible.
Digital image processing refers to processing a digital image by means of digital computer, and the study of algorithms for their transforms. Many of the image processing techniques for monochrome images can extend to colour image by processing the three components image individually. There are many contrast enhancement methods which have been proposed in the literature. Linear contrast enhancement also known as contrast stretching, the original image can linearly expand into a new distribution. The total range of sensitivity of the digital device can be utilized by expanding the original value of an image. This method of enhancement can be mostly used in remote sensing. Non linear contrast enhancement involves the histogram equalization method. The limitation of non linear contrast enhancement is that in which each value of an input image have several values in the output image due to this the original object lose their accurate brightness. Very popular technique for image enhancement is Histogram Equalization (HE). Histogram Equalization (HE) is a method that is used to enhance the contrast of an image. In this it is not necessary that the contrast of an image will always be increased. In some cases, the histogram equalization shows that it can be worse than the contrast of an image is decreased. Before processing histogram equalization, it is necessary to understand the two concepts that are called as PMF (Probability Mass Function), and CDF (Cumulative Distributive Function). For all the pixels in an image, firstly calculate PMF and CDF, then work further. Contrast enhancement tunes the intensity of each pixels magnitude based on its nearby pixels. But it is also known that enhancing the light or value component may not produce effective results in all the cases because change in brightness or light can only change the sharpness of the image. The survey has found that most of the existing methods based on the transform domain techniques, may introduce the artefacts. Transform Domain Method may decrease the intensity of the remote sensing input image. The use of adaptive histogram equalization has been ignored by many researchers to reduce the problem of poor Brightness which will be presented in the output image due to dominant brightness level analysis.
Mathematical morphology is a relatively new approach to image processing and analysis [8-10]. The top hat transformation is used to improve the contrast of the images based on the shape and size of the structuring element [11]. The application of mathematical morphology to image processing and analysis has initiated a new approach for solving a number of problems in the related field [12-14]. This approach is based on set theoretic concepts of shape. In morphology, the objects present in an image are treated as sets. The identification of objects and object features through their shape makes mathematical morphology become an obvious approach for various machine vision and recognition processes. In morphology, the objects in an image are considered as set of points and operations are defined between two sets: the object and the Structuring Element (SE). Basic morphological operations are erosion and dilation. Other operations like opening (closing) are sequential combinations of erosion (dilation) and dilation (erosion) operations. Dilation adds pixels to the boundaries of objects in an image, while erosion removes pixels on object boundaries. The number of pixels added or removed from the objects in an image depends on the size and shape of the structuring element used to process the image. A structuring element is a matrix consisting of only 0's and 1's that can have any arbitrary shape and size. The pixels with values of 1 define the neighbourhood. The top hat transformation provides an excellent tool for extracting bright or dark features smaller than a given size from an uneven background. There are two variations of top hat transformation: white top hat and black top hat transformation. The white top hat transformation helps to extract the white or bright features of the image smaller than the size of the structuring element. The black top hat transformation is used to extract the black or dark features of the image. The white top hat transform relies on the fact that by grayscale opening, one can remove from an image the brighter areas, i.e. features that cannot hold the structuring element. Subtracting the opened image from the original one produces an image where the features that have been removed by opening are clearly visible. Similar thing is true for closing operation also. It means that using a closing operation instead of an opening and subtracting the original image from the closed one helps to extract dark features from a brighter background. This is known as black top-hat transformation opposed to white top-hat transformation in case of opening.
The opening and closing operations are defined, respectively, as:
Where, f is the image, g is the structuring element, ○ denotes the opening operation, and ● denotes the closing operation.
The White Top-hat Transformation (WTH) is defined as the residue between the original image and its opening.
The Black Top-hat Transformation (BTH) is the residue between the closing of an image and the original image.
Where f(x, y) is a gray scale image, and g is a structuring element. Top-hat and bottom-hat transforms which are in gray scale morphology are used for image contrast enhancement as shown in Figure 1.
Figure 1. Steps for Proposed Image Enhancement Technique
The proposed method can be summarized in the following steps.
The computer simulated results have been obtained by implementing the techniques on a set of medical images, and the results have been represented visually. The various techniques being implemented are Histogram Equalization, Contrast-Limited Adaptive Histogram Equalization, linear contrast stretch, image equalized by Gamma, and Multiscale Morphological Filtering. Figures 2, 3, and 4 represent the visual results of implementation of the techniques using MATLAB 7.0.1. In the multi scale morphological approach, “disk” shaped structuring element has been used and the scale factor being taken is “5” for simulation. The Experimental results demonstrate that it improves the visibility and perceptibility of images. Figures 2-4 show that low contrast images are enhanced by using the proposed technique. The quantitative results are compared with standard existing techniques, and the superiority of the proposed technique are justified as tabulated in Tables 1-3.
Figure 2. Contrast Enhancement of Medical Image, (a) Original Image, (b) Result of CLAHE, (c) Result of Histogram Equalization, (d) Result of Linear Contrast Stretch, (e) Result of Gamma, (f) Multi Scale Morphological Operation
Figure 3. Contrast Enhancement of Satellite Image, (a) Original Image, (b) Result of CLAHE, (c) Result of Histogram Equalization, (d) Result of Linear Contrast Stretch, (e) Result of Gamma, (f) Multi Scale Morphological Operation
Figure 4. Contrast Enhancement of Standard Image, (a) Original Image, (b) Result of CLAHE, (c) Result of Histogram Equalization, (d) Result of Linear Contrast Stretch, (e) Result of Gamma, (f) Multi Scale Morphological Operation
Table 1. Qualitative Results for Contrast Enhancement for Figure 2 Proposed Technique Compared with Conventional and Some State-of-Art Techniques
Table 2. Qualitative Results for Contrast Enhancement for Figure 3 Proposed Technique Compared with Conventional and Some State-of-Art Techniques
Table 3. Qualitative Results for Contrast Enhancement for Figure 4 Proposed Technique Compared with Conventional and Some State-of-Art Techniques
The proposed technique has been tested on several different images. In order to show the superiority of the proposed method over the conventional and state-of-art techniques from visual point of view are included in Figures 2-4. In those figures with low-contrast images, the enhanced images by using adaptive histogram equalization, histogram equalization, Local contrast stretch, Equalized by Gamma and also the enhanced images obtained by the proposed technique are shown. It is clear that the resultant image, enhanced by using the proposed technique is sharper than the other techniques.
In this paper, analysis of various contrast enhancement techniques is presented along with their visual and quality comparison. Many image contrast enhancement techniques like Histogram Equalization (HE), Contrast Limited Adaptive Histogram Equalization (CLAHE), and Morphological operation enhancement have been reviewed and compared. The applicability of these techniques in various application domains is also discussed. Experimental results are obtained on various medical, satellite, and standard images including a synthetic image. The morphological operations based on the enhancement approach provides good results in comparison to the results obtained with other state-of-art techniques.