Implementation of a Smart Ward System in a Hi-Tech Hospital by Using a Kinect Sensor Camera

Abstract

With the evolution of sophisticated image capturing device, the potential charecteristics of image processing are being harnessed in various commercial applications. This paper presents a unique idea of using image processing in the field of health monitoring system. Research have been carried out in communication and networking for the design of Smart Ward Systems used in Health Monitoring System that empowers hospital staff to focus more on patient’s dynamic movement by enabling the collection of data at the bedside and remove the need for duplication and double-handling. The proposed work is designed for detecting and recognizing the gesture obtained through the Kinect camera. The training includes data collection and feature extraction. The trained data is then classified using k-nearest neighbors, support vector machines and artificial neural networks. To adopt the best classifier, this paper compares the accuracy of all the above techniques. Mode selection operation has been tested with three different classifiers and support vector machine is proven to be best out of them. The evaluation is done based on various performances metrics like classification effectiveness, accuracy and recognition rate.