Purchase this article with an account.
Ann Judel Enriquez, George J Andersen, Craig W Sauer; Static scene analysis for the perception of heading: landmark identity and position information. Journal of Vision 2003;3(9):131. doi: https://doi.org/10.1167/3.9.131.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recently we have shown that observers use a scene based analysis for the perception of heading in the absence of apparent motion (Hahn, Andersen, & Saidpour, in press, Psychological Science). In the following experiments we investigated the importance of objects as landmarks in performing heading judgments under these conditions and whether the information encoded is based on the object identity or the positions of the objects. 2 frame sequences of computer-generated scenes were presented that simulated a change in observer position through the scene. Participants judged whether they moved left or right from the initial position depicted on the first frame. Scenes were presented with either a short inter-stimulus interval (ISI) of 50 ms to maximize apparent motion, or a long ISI of 1000 ms to eliminate apparent motion. In Experiment 1 we examined whether accuracy increased when a textured scene contained objects compared with textured scenes without objects. Accuracy increased with objects present in the scene, indicating that observers use objects as landmarks to infer observer motion. In experiment 2 we examined whether observers used object identity by preserving the locations of objects across frames but changing the object identity on half the trials. Greater accuracy in direction judgments occurred when object identity was not changed, suggesting that observers encode object identity to perform the task. Surprisingly, accuracy was above chance levels when object identity changed across frames, suggesting that observers also encoded position information of the objects independently of their identities. We conclude that objects serve as landmarks in determining heading, and that participants use both object location and identity when inferring the path of motion from static scenes.
This PDF is available to Subscribers Only