The Application of Image Processing to Determine Lift Occupancy

Abstract

Abstract Lifts have become an essential device to move people up and down tall buildings every day. However, sometimes the lift still stops even though it is already fully occupied, either with people or objects, and is not able to take in more passengers. This makes the amount of time used to get to a floor longer due to all the unnecessary delays. As a result, the primary emphasis of this research endeavour is placed on the application of image processing techniques, namely those that make use of the tracking system that uses Kalman filter, to track moving people and objects, to determine the occupancy of a lift. Kalman filter is applied to accurately track and determine whether it is a person or an object that is located within the bounding boxes. If it is a person, an estimate of the amount of space that an average human occupies in a given environment is determined. In contrast, if it is not a human, then the usage of the polygon function will be incorporated to estimate the percentage of space that is occupied by the items. This will be done in order to figure out how much space the objects take up. Through the use of the proposed system, the occupancy of the lift is determined by analyzing still images taken from a video that was recorded by a camera located within the lift. As a consequence of this, the average travel time for a lift user can be reduced with the reduction of unnecessary stops while it is fully occupied.