image

Same Spot Diff

Align slightly mismatched before/after photos and highlight the areas that truly changed.

Local processing

About your filesFiles and content selected in this tool are processed in your browser and are not uploaded to a Browser Kitty backend.

Operation demo See the basic workflow in a short video.

About this tool

Same Spot Diff compares Before and After photos even when they were not taken from exactly the same position. It uses visual features to compensate for translation, rotation, and mild perspective differences before detecting change, reducing the false impression that the whole image changed because of camera movement.

It is designed for construction records, equipment or facility checks, store displays, shelves, walls, panels, and other repeat-photo inspections. An optional camera guide overlays the Before image on the live view to help capture a better-aligned After photo.

Good for

Compare before-and-after construction or repair photos to find what actually changed
Inspect equipment, facilities, panels, or shelves from recurring photos
Check which parts of a store display or signage changed since the previous photo
Use the Before overlay as a guide when taking a later After photo from a similar viewpoint

What it can do

Key capabilities available in this tool.

Automatic alignment for position, rotation, and mild perspective differences
Highlight changed areas in red
After-camera guide overlaid with the Before image
Slider, blink, and overlay comparison modes
Exclude user-defined regions from comparison

How to use

Open the tool and add the Before photo.
Add an After photo, or use the camera guide to overlay Before on the live view while taking the After image.
Run the comparison and start with the Change view to inspect red-highlighted regions.
Use Alignment, Slider, or Blink views as needed to visually check residual alignment and the Before/After difference.
Exclude regions that change every time, such as clocks, reflections, screens, or moving leaves, then adjust the result if necessary and save the current view as PNG.

Supported

Input images: JPEG, PNG, and WebP. The tool compares one Before image with one After image.
Alignment uses ORB features, BFMatcher, RANSAC homography, and perspective warping to compensate for camera position, rotation, and mild perspective differences.
Views include Change, Alignment, Slider, Blink, and individual source images, with pinch zoom and drag support on mobile.
The After camera guide requires camera permission and is most reliable over HTTPS. The current comparison view can be saved as PNG.

Limitations and notes

If the viewpoint changes substantially, foreground and background parallax can differ too much for a single homography to align the entire scene accurately.
Alignment may fail on nearly textureless scenes, heavily occluded photos, or images whose overall composition changed significantly.
Strong shadows, lighting changes, reflections, moving leaves, screens, and clocks can be detected as changes even when the underlying object did not change. Use adjustment and exclusion regions when appropriate.
Automatic detection is an aid to visual inspection. For important construction, maintenance, or safety decisions, also review the original images and alignment quality.

Frequently asked questions

Do the two photos need to be taken from exactly the same position?

No. The app can compensate for position, rotation, and mild perspective differences, but large viewpoint changes or strong parallax may exceed what the alignment can correct.

Are the photos uploaded to Browser Kitty?

No. Alignment, difference detection, the camera guide, viewing, and PNG export run in the browser; selected photos are not uploaded to a Browser Kitty backend.

Can I exclude a clock, monitor, or other area that changes every time?

Yes. Add exclusion regions on the comparison result. You can hide the guide outlines while keeping the exclusions active.

Can it help me take the After photo from a similar viewpoint?

Yes. The After camera guide overlays the Before image on the live camera view. Camera availability depends on HTTPS context and browser permissions.

Offline version and source code

The standalone same-spot-diff.html bundles the required OpenCV JavaScript and WebAssembly. Normal photo comparison can run offline from the saved HTML; the After camera guide depends on browser camera permissions and secure-context requirements.

View source on GitHub