Skeleton Based Human Action Recognition Using Doubly Linked List

Muhammad Sajid Khan, Andrew Ware, Usman Habib, Muhammad Junaid Khalid, Nisar Bahoo

    Research output: Contribution to journalArticlepeer-review

    33 Downloads (Pure)

    Abstract

    Human Action Recognition is a significant focus for research because of its many applications in robotics and automation. This paper demonstrates how doubly linked lists can be used to sequence the 3D human actions recorded as video clips in the NTU RGBD 60 dataset. The nodes and edges in the list represent the joints and bone structure in the human skeleton. Each node holds information about the joint’s position within the skeleton and pointers to its parent and child nodes. The doubly link list is constructed by first utilising the nodes representing the torso joints and then adding the nodes for the limbs’ joints. The chosen sequence of nodes preserves the structural shape of the skeleton. The linked lists for many known activities are used as the training set for a classifier capable of identifying subsequent human actions. The classifier is based on the displacement between consecutive nodes in the action sequence. This approach minimises the complexity of the tree structure and improves the accuracy of 3D action recognition.
    Original languageEnglish
    Pages (from-to)18-21
    Number of pages4
    JournalInternational Journal of Computer Trends and Technology
    Volume70
    Issue number2
    DOIs
    Publication statusPublished - 24 Feb 2022

    Keywords

    • Skeleton based action recognition
    • Human Action recognition
    • video processing
    • doubly linked list

    Fingerprint

    Dive into the research topics of 'Skeleton Based Human Action Recognition Using Doubly Linked List'. Together they form a unique fingerprint.

    Cite this