Pixel Mask
Generate a binary mask image where changed pixels are white and unchanged pixels are black. Useful as an alpha mask or for further programmatic processing.
Click or drag image here
PNG, JPG, WebP, GIF
Click or drag image here
PNG, JPG, WebP, GIF
What is a Pixel Diff Mask?
A pixel diff mask is a greyscale image where each pixel is set to pure white if the corresponding pixel in Image A differs from Image B by more than the tolerance, and pure black otherwise. The result is a binary image - no intermediate grey values.
Masks are designed for programmatic use: you can load the PNG into an image editing application or pipeline and use it as a selection, an alpha channel, or an input to further processing such as dilation, erosion, or connected-component analysis.
Inverted Mask
When "Invert mask" is checked, the colours are flipped: unchanged pixels become white and changed pixels become black. An inverted mask is useful when you want to select the stable areas of an image rather than the changed areas.
Common Use Cases
Compositing and Blending
Use the mask as an alpha channel to blend only the changed regions of Image B over Image A, creating seamless composites.
Automated Test Gating
Feed the mask into a CI pipeline to count white pixels. If the count exceeds a threshold, reject the build due to visual regression.
Change Map Export
Export the mask alongside the original images as documentation of exactly which regions changed between two versions of a design.
Image Segmentation Prep
Use movement or change masks as training labels or annotations for machine learning models that perform segmentation or object detection.