Image contrast enhancement using Histogram equalization with Fuzzy Approach on the Neighbourhood Metrics (FANMHE)

In this paper, a fuzzy approach on the neighbourhood metric is proposed for usage with Histogram equalization for image contrast enhancement. The Neighbourhood Metrics or local image properties are used to sub divide the large histogram bins produced by Global Histogram Equalization. Large histogram bins in the image causes visual deteriorations. We propose, a fuzzy approach to sub divide the large bins. Fuzzy Approach on Neighbourhood Metric for Histogram Equalization(FANMHE) consists of two stages. In the first stage, Gaussian membership function is used to a assign membership value for all the eight neighbouring pixels with reference to the middle pixel. The next stage Fuzzy C-Mean clustering algorithm is used to group the pixels with similar membership value. After clustering pixels with similar neighbourhood histogram equalization performed. This approach sub divides large histograms bins into sub bins giving equal weight age to all the 8 neighbouring pixels. The pixels with similar neighbouring properties considering 8 pixels are grouped into each sub bins. This provides actual and meaning grouping to divide large histogram bins into sub bins.