Evaluating 2D Flow Visualization Using Eye Tracking


(a)      diff.= 11.18

(b)      diff. = 66.37

(c)       diff.=77.66

(d)      diff.=59.77

(e)      diff.=127.56

Five eye scanpathes and their trajectory difference with the pathline in Exp 2: Advection Prediction. (a) Eye gaze moved with the pathline. (b) Eye gaze kept scanning on the pathline repeatedly. (c) Participant’s eye gaze moved to a distant location. (d)(e) Participant focused on arrows first and traced the pathline later. Please see also the supplemental materials for enlarged images.




Flow visualization is recognized as an essential tool for many scientific research fields and different visualization approaches are proposed. Several studies are also conducted to evaluate their effectiveness but these studies rarely examine the performance from the perspective of visual perception. In this paper, we aim at exploring how users’ visual perception is influenced by different 2D flow visualization methods. An eye tracker is used to analyze users’ visual behaviors when they perform the free viewing, advection prediction, flow feature detection, and flow feature identification tasks on the flow field images generated by different visualizations methods. We evaluate the illustration capability of five representative visualization algorithms. Our results show that the eye-tracking-based evaluation provides more insights to quantitatively analyze the effectiveness of these visualization methods.




l       Hsin-Yang Ho, I-Cheng Yeh, Yu-Chi Lai, Wen-Chieh Lin, and Fu-Yin Cherng, “Evaluating Evaluating 2D Flow Visualization Using Eye Tracking,EuroVis 2015 (Computer Graphics Forum, Vol. 34, No. 3).


Additional results


        Supplemental materials