A hybrid cluster technique for improving the efficiency of colour image segmentation. (25th November 2020)

• Record Type: Journal Article
• Authors: Kumar, Ankit; Raja, Linesh; Dadheech, Pankaj; Bhardwaj, Manish
• Is Part Of: World review of entrepreneurship, management and sustainable development. Volume 16:Number 6(2020)
• Page Start: 665
• Page End: 679
• Publication Date: 2020-11-25
• Subjects: data mining -- image processing -- classification -- image segmentation -- clustering -- k-mean cluster; Sustainable development -- Periodicals; Entrepreneurship -- Periodicals; 338.04
• Languages: English
• ISSNs: 1746-0573

Abstract

Image segmentation is a wide area for researching, and in many applications, segmentation is applied for finding the distinct group in the feature space. It separates the data into different regions or clusters, and each one is homogeneous. The current algorithm, which is proposed approach for noise reduction, eliminates most of the noise from the input image. This noise concerns to cut boundary of the noise full image. The result shows that it is very efficient in segmenting the image and reduces the time complexity. The proposed algorithm can be used in object deduction or in object analysing in image processing. The segmentation of image proceeds by using combination of different segmenting approach in 3-D RGB colour-space. Clarity of the output segmented image is better in comparison to other segmentation techniques of the image. Clarity of the output image is depending on the number of clusters used.