TY - JOUR
T1 - Use of interreflection and shadow for surface contact
AU - Madison, Cindee
AU - Thompson, William
AU - Kersten, Daniel
AU - Shirley, Peter
AU - Smits, Brian
PY - 2001/2
Y1 - 2001/2
N2 - The interaction of light with surfaces results in a number of lighting effects that may serve as valuable visual cues. Previous research on shadows has shown them to be effective in determining the three-dimensional (3-D) layout of a scene, but interreflections have been ignored as cues for spatial layout. Interreflections as well as shadows may help to disambiguate the 3-D layout of objects by providing information about an object's contact with a surface. We generated computer images of a box on an extended textured ground plane that was either in contact with the ground or was slightly above the ground. Images were rendered for four conditions: (1) no shadow + no interreflection, (2) shadow only, (3) interreflection only, and (4) shadow + interreflection. A photometrically incorrect condition was also included. The participants rated the degree of contact for each image on a scale, which was used to generate receiver operating characteristic (ROC) curves and a measure of sensitivity. In the images with no shadows or interreflections, the participants performed at chance levels. Interreflections, shadows, and a combination of interreflections and shadows all resulted in high sensitivity for judging object contact. More important, information from shadows and interreflections can be combined, resulting in near-perfect judgment of surface contact. Interreflections and shadows can be effective cues for object contact.
AB - The interaction of light with surfaces results in a number of lighting effects that may serve as valuable visual cues. Previous research on shadows has shown them to be effective in determining the three-dimensional (3-D) layout of a scene, but interreflections have been ignored as cues for spatial layout. Interreflections as well as shadows may help to disambiguate the 3-D layout of objects by providing information about an object's contact with a surface. We generated computer images of a box on an extended textured ground plane that was either in contact with the ground or was slightly above the ground. Images were rendered for four conditions: (1) no shadow + no interreflection, (2) shadow only, (3) interreflection only, and (4) shadow + interreflection. A photometrically incorrect condition was also included. The participants rated the degree of contact for each image on a scale, which was used to generate receiver operating characteristic (ROC) curves and a measure of sensitivity. In the images with no shadows or interreflections, the participants performed at chance levels. Interreflections, shadows, and a combination of interreflections and shadows all resulted in high sensitivity for judging object contact. More important, information from shadows and interreflections can be combined, resulting in near-perfect judgment of surface contact. Interreflections and shadows can be effective cues for object contact.
UR - http://www.scopus.com/inward/record.url?scp=0035256975&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0035256975&partnerID=8YFLogxK
U2 - 10.3758/BF03194461
DO - 10.3758/BF03194461
M3 - Article
C2 - 11281095
AN - SCOPUS:0035256975
SN - 0031-5117
VL - 63
SP - 187
EP - 194
JO - Perception and Psychophysics
JF - Perception and Psychophysics
IS - 2
ER -