Skip to content
opencv-wasm

Image resizing

Algorithm atlasGeometry

Sample the source on a new pixel grid at a different resolution.

Try it on your image ↓

VISUAL WALKTHROUGHGeometry
REFERENCE INPUTFour neighbouring source samples

A magnified patch makes the interpolation arithmetic visible.

RESULTInterpolate one destination value

Repeat the mapping and interpolation for every destination pixel. Other interpolation flags use different sampling rules.

STEP 03 / 03

Combine the samples into one output pixel.

Illustrative example. The stages explain the method; they are not a live OpenCV execution.

Try it on an image

YOUR IMAGE · REAL OPENCV

Experiment at pixel level

Open full lab ↗

Resample the colour image. Input and output have different pixel coordinates when their sizes differ.

The engine loads on your first run. Your images stay in this browser.

Input448 × 320
OutputWaiting for a result

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
Hover to inspect. Click to pin a pixel.
Input 
Select a pixel

Output 
Select a pixel

Sample models and licenses

When to use it

Create thumbnails, image pyramids or model inputs.

How it works

  1. 01Map each destination pixel back to a source location.
  2. 02Choose neighbouring source samples using the interpolation rule.
  3. 03Combine the samples into one output pixel.

destination grid → source coordinates → interpolation

What to tune

INTER_AREA is useful for shrinking; nearest preserves discrete labels; linear and cubic trade sharpness against cost.

Where it breaks down

Downsampling can alias high-frequency texture. Premultiply RGB before resizing transparent images to avoid coloured fringes.

TypeScript API

Open an entry for its exact overloads, parameter descriptions, result ownership and pinned upstream source.

resizeReference →

All of these calls execute on the CPU. Native objects need explicit disposal. See matrices and ownership and build compatibility.

Arbitrary remapping

Provide a source coordinate for every destination pixel.

Affine transformations

Map points with a linear transform plus translation, preserving straight lines and parallelism.

Homography and RANSAC

Estimate a projective mapping between two views of a plane, rejecting inconsistent point matches.

Polar and log-polar warps

Re-express an image using angle and radius around a chosen centre.