Contents I. Introduction
Digital imagery and its various kinds of indispensable contributions in various anthropological applications ranging from handhold capturing devices, infrared-imaging, biomedical imaging, radar imaging, SAR imaging, sonar imaging to remotely sense imaging systems, are still struggling for subjective contrast as well as entropy content enhancement as a pre-processing stage. Hardware limited image procurement ability and unstable brilliance, collectively lead to dark and low contrast images and in such kind of images average intensity strength is much diminished. Hence, quality improvement and information restoration (textural as well as edge details along with acceptable intensity spread) is an indispensable ingredient of image processing. A plethora of literature is already available for quality enhancement purposes as various methodologies are suggested using spatial domain, histogram domain as well as frequency domain operations. Here, in this work fuzzy based or fuzzified histogram redistribution is suggested using sigmoidal membership function so that its cumulative distribution can be utilized for gamma value-set evaluation and later, its cosine transformed coefficients’ energy redistribution is suggested for textural boosting. Various enhancement approaches are discussed in the chronological order in an organized manner in [1 –5]. Histogram-based processing [3] either in global as well as the local manner is highly appreciated and hence, various kinds of histogram sub-division approach usually followed by the corresponding sub-equalization have been suggested. In this context, Recursive mean-separate HE (RMSHE) [6] was suggested by following the recursive separation principle. A sequel of this algorithm termed as recursive sub-image HE (RSIHE) [7] was also proposed. Here, the proposal is to segment the intensity span using median intensity level, rather than mean intensity value that was earlier suggested for RMSHE. By this time, researchers started looking up for various ways of histogram sub-division depending upon various thresholds in one manner or the other. In this sequence, for imparting more adaptive behavior, another approach is suggested, which is termed as median-mean based sub-image-clipped HE (MMSICHE) [8]. In this method, adaptive histogram-clipping is suggested along with median based intensity span thresholding which itself is followed by further segmentation of both of these bins using their individual mean values. Finally, histogram sub-equalization is employed as an indispensable ingredient in such kind of methodologies, which leads to a similar kind of short-comings of false intensity re-allocation sometimes. Fuzzy intelligence has been also introduced in image enhancement in a form of significant contribution like Brightness preserving fuzzy dynamic HE (BPFDHE) [9]; which itself is a kind of extension work framing the backbone of dynamic HE (DHE). Due to the employed fuzzification, a kind of highly adaptive histogram smoothening has been suggested, but for sufficient boosting up of intensity values especially while dealing with dark images, there is no any provision and hence, the need for gamma-correction-like approaches came to the existence. In the same sequence, the averaging HE (AVGHEQ) [10], RHE-DCT [11] and later onwards, HE based optimal profile compression (HEOPC) [12] has been proposed for color image enhancement. Focusing on allowable intensity span expansion through count reduction of void intensity bins for color image enhancement has been proposed under the head of the histogram equalization with maximum intensity coverage (HEMIC) [13]. For general images, these techniques seem sound; but without imparting gamma correction [14–15], it is very tough to impart quality improvement for dark images. Earlier, the adaptive gamma correction with weighting distribution (AGCWD) [16] is introduced, but sometimes leads to the regional over-saturation due to nature of its transformation curve and hence, improved versions [17–18] also came to the existence. Other better proposals have been also drafted like the intensity and edge-based adaptive unsharp masking filter (IEUMF) [19] by employing the unsharp masking filter which can be further utilized as edge augmentation for imparting overall enhancement. Rest of the paper is drafted as follows: Section II deals with the proposed fuzzy derived dynamic equalization based gamma corrected DCT based energy redistribution for color image enhancement. Performance evaluation and comparison based Experimental results are presented in Section III and finally, conclusions are drawn in Section IV.
