Performance evaluation of segmenting images using improved FELICM based gray stretch

Segmenting an image plays a crucial role in digital image processing. Segmentation depicts how the object is composed in an image and aids in extracting useful information from it so that interpretation becomes easier. Segmentation may be regarded as an essential pre-processing stage that will help to acquire objects that can be efficiently utilized in further steps of processing. This extraction of objects is helpful in almost every field today such as medical, forensic, industry and agriculture. Instead of using pixels, DWT make use of waves that helps to extract object efficiently. To raise the segmentation accuracy further an improved Fuzzy-C Mean with Edge and Local Information (FELICM) based gray stretch strategy is suggested that will help to segment an image by obtaining an optimistic and adaptive threshold by incorporating FELICM and gray stretch on coefficients with low frequency and edge enhancement on coefficients with high frequency after DWT. Existing and the proposed strategies have been compared using parameters Accuracy, Uniformity, Shape measurement and Regional Contrast.