Analysis and Identification of Human Motor Activity using KNN algorithm

1K Venkata sumanth, Shobana priya

119 Views
43 Downloads
Abstract:

This paper proposes a method for human activity detection using K-Nearest Neighbour algorithm. By using this algorithm basic activities of the human are walking, running, hand waving, hand clapping and jogging is detected. Initially video is comprised into frames Next feature extraction takes place. Feature extraction is filtering the high pass, low pass noises and detecting the corners of the human poster in every frame. This is done using Harris corner detection algorithm. Next by using corner points in every frame the activity of the human poster is detected using K-Nearest Neighbour algorithm. In this Kth dataset is used for testing the algorithm.

Keywords:

K- Nearest neighbour algorithm, CNN algorithm, Harris corner detection, Human activity detection, human motor identification

Paper Details
Month4
Year2020
Volume24
IssueIssue 4
Pages8726-8732