A Study on Parallax Error in Video Analysis

Abstract

The use of video analysis for the study of motion and forces is well established in the worlds of physics and biomechanics. The issue of parallax error, sometimes referred to as perspective error or scaling error, when videotaping a moving object is known, and the experimenter is often counseled to be “far away.” However, “far away” is not often defined and sometimes experiments have limitations on the distances between camera, object, and scale. Therefore, the purpose of this paper is to look into how far away is far enough. Our paper is not meant to be an exhaustive look at errors in video analysis; rather, we hope it may find use as a simple guideline when setting up your own video analysis experiments. We approach the error from parallax in two ways, first as a relative error within the data, and second via an appendix, as a correction factor based on experimental setup. We offer this work as a guide and perhaps a rule of thumb for those using video analysis as a tool in their research or their classroom. We also suggest this experiment as a possible entry point to the topic of experimental errors.