lec5 4pp

lec5 4pp - Lecture 5: Sept 8, 10 Last Class Brightness and...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
USC CS574: Computer Vision, Fall 2010 Copyright 2010, by R. Nevatia 1 Lecture 5: Sept 8, 10 • Last Class – Brightness and color • Today’s objective – Color Constancy – Image Segmentation USC CS574: Computer Vision, Fall 2010 Copyright 2010, by R. Nevatia 2 Uses of Color • Spectral albedo is characteristic of surface properties – Helps in segmentation as different surfaces are expected to have different “colors” – Helps distinguish material properties •R ipe vs raw fruit • Slippery vs dry ground conditions • …. . • To be effective for these tasks, we need to be able to infer surface spectral albedo even though the measurements combine color of illumination, surface and sensor responses (the last one could be calibrated). USC CS574: Computer Vision, Fall 2010 Copyright 2010, by R. Nevatia 3 Color/Brightness Constancy • Perceived color invariant to illumination (to some extent) – Note difference in illumination spectral distribution between natural light and tungsten filament light Lightness (perceived brightness) of a surface also invariant – Consider taking a white sheet of paper from a room to outdoors –I r i s accommodation explains only a small part • Simultaneous Contrast (an example) • Color (lightness) constancy theories – Not well developed, not studied in this course USC CS574: Computer Vision, Fall 2010
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
USC CS574: Computer Vision, Fall 2010 Copyright 2010, by R. Nevatia 5 Simultaneous Contrast • Perceived brightness is also strongly influenced by contour information (examples from E. Adelson, MIT) USC CS574: Computer Vision, Fall 2010 Copyright 2010, by R. Nevatia 6 Some steps in Image/Scene Analysis • Infer a 3-D map of the environment (for obstacle avoidance while moving) • Detect (segment) objects in the environment • Recognize objects and their relations • Observe object motion and infer activity in a sequence of images • Above steps are inter-connected • Issues
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 6

lec5 4pp - Lecture 5: Sept 8, 10 Last Class Brightness and...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online