CAMERA AND SMARTPHONE BASED AUTOMATIC UNUSUAL EVENT DETECTION SYSTEM IN INDOOR ENVIRONMENT

Unusual events which happen at home like fainting have been a serious soundness exposure that decrease the enjoyment of life among society. It leads to implement a home based automated monitoring system like unusual or fall event detectors. A camera integrating a special algorithm was used to classify user’s behavior and smartphone’s accelerometer sensor was used to gather acceleration samples. After completing implementation of the project we did an experiment on our project. From experimental results we can see that the accuracy of the given algorithm was 91.4%, the precision was 91.6%, and the recall was 98%, demonstrating desired achievement of our approach in finding unusual situations.