Skip to content
Free public beta until December 4, 2026: every course is open, no card needed. After that, the first chapter of each course stays free and full access is a one-time $39 ($19 for up to 100 founding members chosen for helpful feedback). Your progress is kept. How the beta works

corner (image-feature sense)

A location where the local image structure constrains position in every direction at once, not just one.

A corner is a pixel location where the windowed sum-of-squared-differences function E(u,v)=\sum_{(x,y)\in W} w(x,y)\big[I(x+u,y+v)-I(x,y)\big]^2 is large for a shift (u,v) in EVERY direction, not just one. A flat region has E small in every direction; an edge has E small along the edge's own direction (the aperture problem: a window sliding along a straight boundary looks unchanged) and large only across it. A corner is the one case with no such ambiguous direction, which is exactly what makes it usable as a well-localized point to detect and later match between two images.

Read the full section →
Graph settings

Calculator