SMART IMAGE TOOLS IMAGE WORKSPACE
A simple background. A transparent PNG. Control every step.
Pixels are inspected and edited on your device. No image upload or AI model download.
Drag and drop a file here, or choose from your device.
Original synthetic practice image · created on your deviceJPG, PNG or WebP · up to 50 MB and 8 megapixels · still-image processing
METHOD & PRACTICAL GUIDANCE
The first color starts from the top-left pixel. Click another opaque background pixel or enter its coordinates to sample a better match. For a painted checkerboard, enable the second color and sample its other tone. Start with low tolerance; higher values can erase subject colors close to the selected backgrounds.
Edge-connected mode follows matching and already transparent pixels inward from the image boundary. A white detail fully enclosed by a dark logo can remain. All matching pixels also removes enclosed matching regions, which can help with spaces inside letters but can erase intentional details. Transition softness introduces partial transparency in colors just beyond the tolerance threshold.
Apply the settings and compare the source with the cutout at 100% zoom. Check on light and dark backdrops for leftover checkerboard fragments or pale halos. The alpha counts report what is actually transparent. Changing a setting clears the old result so the next PNG uses the settings you have reviewed. Downloads retain the original dimensions.
This is deterministic color matching for simple graphics. It cannot reconstruct foreground colors already blended into an opaque background. Similar colors at a subject boundary may be removed; use the AI remover for varied photographic backgrounds. How we document our tools
A PRACTICAL EXAMPLE
Clear a simple two-tone background while protecting enclosed details.
Similar foreground colors connected to the background can be removed too. Softness changes alpha but does not reconstruct foreground colors lost to a flattened background.
Examples illustrate specific inputs and settings. Actual output size and appearance depend on your file and browser. Practice images are synthetic; they do not contain personal data.
GOOD TO KNOW
For a simple checkerboard with two tones, enable the second color and sample each tone. Edge-connected removal removes matching regions reachable from the image border. JPEG artifacts or antialiasing may need tolerance adjustments, and similar subject colors can also be affected. It is not automatic pattern recognition.
It only edits matching pixels connected through matching or already transparent pixels to an image edge. Enclosed matching details can remain. All matching pixels also edits enclosed regions, such as a background inside a letter, and may remove matching subject details.
It is best for logos, graphics and simple backgrounds. Use the AI background remover for varied photographic backgrounds. Color removal cannot recover original foreground colors from a flattened image, so inspect fine edges and possible halos.
The cutout downloads as PNG with the original pixel dimensions. The JSON report includes transparency counts. JPG cannot preserve transparent pixels.
No. Pixel analysis and editing run in a worker on your device. No account or AI model download is required. JPG, PNG and WebP are supported up to 50 MB, 8 megapixels and 8,192 pixels per side. Animated images are inspected as a decoded still frame.