prkaritt / har

HAR

The relationship between human and computer is deepening day by day. Computers have made life of human beings easier and comfortable in many ways. The recognition of activity performed by humans by computer has gained a lot of popularity in robotics and computer vision field. This recognition can open the door and widen the opportunities for a lot of applications. Inclining to this growing popularity, interest and opportunities that this field has to offer, we need to build models that recognizes the different kinds of activities that humans perform. So, in this project we have selected six major activities that are performed by humans and put some efforts to deduce a model that can classify the action that an individual is performing based on the accelerometer data that is collected while the action is being performed. The six activities that can be classified are: Walking, Sitting, Downstairs, Upstairs, Jogging, standing. Deducing this kind of model can server different purposes. It can be used to track the different activities that an individual performs and display them the report. It can also be used in fields like computer animation in order to model the characters. It has many more applications. So, as stated above, the objective of this project is to classify which activity an individual is performing out of those six activities.