Anthropomorphic prosthetic hands must resemble human hands concerning their kinematic abilities and significant features such as object recognition on grasp. Implementation of object recognition by prosthetic hands require prior information on the extrinsic structural properties of objects as well as the hand's position and orientation. Object recognition approaches are primarily of two categories: vision and tactile. Vision-based learning methods have dominated the realm of object recognition for robotic and prosthetic hands in the past decades. Relying only on vision is not sufficient for the perceptual requirements of a prosthetic hand. The human hand is a complex structure with multiple degrees of freedom (DoF), leading to various movements and grasp formations. Acquiring the full range of motion in the fingers and the wrist during prosthetic hand development is critical. Creating such dexterity involves an intense investigation to extract knowledge of motion and joint constraints in the phalanges and wrist bones. The increase of digital information in the current age has elevated the demands of semantically rich annotations for applications shared over the internet. In recent years, the popularity of philosophical knowledge representation methods like ontology to under- stand and utilize relevant domain concepts for problem specifications has escalated. Ontologies are significant in providing a meaningful schema by linking unstructured data. Object recognition during a grasp is an essential attribute in a prosthetic hand, which takes its development closer to its natural counterpart.