An Eye Tracking Based Virtual Tape Measure in Construction

Abstract

The metal tape measure is one of the most important tools used by construction workers on a daily basis. The measure allows the workers to make sure that their piece of work is completed correctly following designs and/or specifications. Therefore, it is a must-have item in the worker’s toolkit. However, existing metal tape measures can be subject to manipulation difficulties, despite that they are durable, retractable, and affordable. A stake, marking flag, or second person is commonly required when measuring large areas. In addition, the metal hook or blade of the measure could injure the worker such as skin cuts. This research proposes a novel framework of virtual tape measures with an eye tracker and a stereo camera. The framework consists of three components: data collection for point-of-interest, sensor calibration and distance calculation. Specifically, the point-of-interest is generated when the worker stares at one location and closes his/her eyes for at least 1 second. Different kinds of data related to the point-of-interest are collected for further analysis. Then, the eye tracking device and stereo camera are calibrated from their captured images to obtain the transformed gaze point on the stereo camera image. Based on the transformed gaze point, the corresponding point cloud is extracted and unified for the distance measurement. The framework was tested by measuring dimensions of several elements. The measurements by the framework were compared with the tape measures to benchmark its accuracy. Results showed that the proposed method could perform the measures efficiently in a hands-free manner with an average absolute error of 2.8 cm.