Detection and separation of generic-shaped objects by fuzzy clustering

Purpose - Existing shape-based fuzzy clustering algorithms are all designed to explicitly segment regular geometrically shaped objects in an image, with the consequence that this restricts their capability to separate arbitrarily shaped objects. The purpose of this paper is to introduce a new detection and separation of generic-shaped object algorithm.Design methodology approach - With the aim of separating arbitrary-shaped objects in an image, this paper presents a new detection and separation of generic-shaped objects (FKG) algorithm that analytically integrates arbitrary shape information into a fuzzy clustering framework, by introducing a shape constraint that preserves the original object shape during iterative scaling.Findings - Both qualitative and numerical empirical results analysis corroborate the improved object segmentation performance achieved by the FKG strategy upon different image types and disparately shaped objects.Originality value - The proposed FKG algorithm can be highly used in applications where object segmentation is necessary. Likewise, this algorithm can be applied in Moving Picture Experts Group-4 for real object segmentation that is already applied in synthetic object segmentation.