Visual saliency
Estimate conspicuous regions or candidate object windows from appearance or motion.
The same scene stays here while the working view changes.
Extract the method’s spectral, local-contrast, motion or gradient features.
The illustrative response concentrates around the smaller, brighter object. The chosen saliency method defines the actual scoring rule.
Thresholding or ranking the saliency response can propose regions for later processing. Saliency itself does not assign object identities.
Return a saliency map or candidate windows for downstream use.
Illustrative example. The stages explain the method; they are not a live OpenCV execution.
Try it on an image
Experiment at pixel level
Compute a static saliency map. Bright values suggest visually distinctive regions, not necessarily recognized objects.
The engine loads on your first run. Your images stay in this browser.
Scroll over either image to zoom at the pointer. Use the scrollbars to pan both views over the same relative area. Zoom is relative to the input; pixel coordinates belong to each image. Warps can change scene correspondence.
Pixel inspector RGBA · native values · matched scale · 9 × 9 output pixels
Select a pixel
Select a pixel
When to use it
Prioritize candidate regions for later processing or explore image attention cues.
How it works
- 01Extract the method’s spectral, local-contrast, motion or gradient features.
- 02Score locations that differ from their surrounding regularities.
- 03Return a saliency map or candidate windows for downstream use.
visual regularity → local exception → candidate attention
What to tune
Select a static, motion or objectness method to match the task. Interpret each output on its own documented scale.
Where it breaks down
Saliency is not an object label, and attractive texture can score more highly than the intended subject. BING needs trained model data.
TypeScript API
Open an entry for its exact overloads, parameter descriptions, result ownership and pinned upstream source.
All of these calls execute on the CPU. Native objects need explicit disposal. See matrices and ownership and build compatibility.