Image Similarity Scanner

Both images are normalized to the same resolution and compared pixel-by-pixel across all three color channels (red, green, blue). The result is a percentage similarity score and a visual heatmap showing exactly where differences exist — highlighted regions indicate edits, crops, retouching, color adjustments, or compression artifacts. The heatmap makes it easy to spot localized changes that would be invisible to the naked eye when viewing images side by side. See our image similarity guide for more comparison techniques.

Find duplicate photos across different exports or album versions — for bulk deduplication of larger sets, try the Duplicate Scanner which uses perceptual hashing to group up to 50 images at once. Verify an image hasn't been tampered with by comparing it to a known original. Check how much quality a compressed version has lost relative to the source file. Detect unauthorized edits in photos submitted for publication. Compare before-and-after retouching to quantify the extent of modifications. Verify that product images across different sales channels are consistent.

The Similarity Scanner compares what images look like — pixel content, color, and structure. The Metadata Comparison tool compares what's inside the files — camera settings, timestamps, GPS, and software history. Use visual comparison when you want to see what changed between two versions of an image. Use metadata comparison when you want to know if two photos came from the same camera or were edited differently.

For best results, compare images that started from the same source. Comparing completely different photos will produce low similarity scores that aren't meaningful. When checking for edits, use the highest-quality version available as the baseline. Heavily compressed images will show compression-related differences even if no intentional edits were made. For file-level integrity verification, use a SHA-256 hash to detect any byte-level changes.