Rehabilitation Using Neighbor-Cluster Based Matching Inducing Artificial Bee Colony Optimization

• Published in: Annual IEEE India Conference pp. 1 - 4
• Main Authors: Saha, Sriparna, Ghosh, Ahona
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2019
• Subjects: artificial bee colony; Clustering algorithms; Convergence; Euclidean distance; Feature extraction; Kinect sensor; neighbor- cluster; Optimization; rehabilitation; Training; wrapper method; Yoga
• ISSN: 2325-9418
• DOI: 10.1109/INDICON47234.2019.9028975

Muscle sprain has become a common reason for doctor consultation nowadays because of our sedentary lifestyle. Yoga helps to increase one's all-around fitness of body and strength of mind. Without or with very less involvement of a trainer, the ability of one to do the exercises himself/herself will motivate the subject more and flexibility of performing those movements will increase the regularity also. The novelty of our proposed work lies in the automatic measurement of posture accuracy. The skeletal forms of postures are detected by Kinect sensor initially, then after extracting the required features, those are matched with the already stored features of the training dataset. During the performance, the patterns of movements are recognized using neighbor-cluster approach. For the feature selection process, the wrapper method using artificial bee colony optimization is implemented. The proposed work is suitable for any type of patient as the system can be accessed at any time of the day in any environment and it outperforms the already present works in this field also.