Background removal for product photos in the browser

Marketplaces want a uniform look: product floating on white or transparent, same margins, no clutter. Shooting on a sweep is ideal, but kitchen-table photos happen. Background removal plus crop gets listings live without a studio budget.

QuickConvert’s Image Crop optional removal runs a model locally. Photos do not upload—a requirement when prototypes are secret.

Set expectations

Hair, glass, and mesh are hard. A white mug on a grey shirt is easy. Removal leaves halos when backlighting was harsh. Re-shoot on a plain sheet beats hours of touch-up for catalog work.

Use diffused daylight or two soft lamps. Put paper or foam board behind the product. Fill the frame so you need less aggressive removal. Slight overexposure on white backdrops simplifies cutouts.

Shoot extra pixels—you will crop. Image Resize down to marketplace minimums after crop saves upload time.

1. Drop the photo into Image Crop.
2. Tighten the box to the product before toggling removal.
3. Download PNG with transparency.
4. Convert to JPG with Image Format Converter only if the platform forbids alpha.

Some platforms demand square canvases. Resize to 2000×2000 with padding in a design tool, or center the PNG on white in the same tool.

Batch reality

This page edits one image at a time. A seasonal launch with fifty SKUs may deserve desktop batch scripts. For one-off listings, interactive crop is enough.

Transparency inflates bytes. Image Compress on a flattened JPG export meets strict limits. Keep the PNG master for future seasonal backgrounds.

Compare JPG vs PNG vs WebP when choosing delivery format for Shopify versus Amazon.

Removal for dust on the backdrop is fine. Removal that hides flaws or includes accessories not sold is not. Marketplace rules vary—read the style guide.

GPS and serial numbers in EXIF may vanish on export. Scrub metadata if leaks matter.

Instruction sheets sometimes need Image OCR for spec tables. Warranty PDFs combine photos via Images to PDF. Watermark internal sell sheets with PDF Watermark.

Yellow indoor bulbs tint edges orange. Removal may leave a warm fringe. Set white balance in the camera app or convert to neutral gray backdrop.

Does the model download every time? First removal may fetch weights; later runs reuse cache.

Can I remove backgrounds on phone? Yes in a modern mobile browser; large images may be slower.

Will JPG keep transparency? No. Use PNG or WebP with alpha.

How do I fix hair halos? Crop tighter, soften backdrop, or shoot again.

Is removal good for people? Portraits are harder than products. Expect manual touch-up for IDs.

Can I batch zip outputs? Convert and compress batches separately; crop is per file here.

What about 360 spins? Use video platforms; this tool is stills.

Does Amazon accept WebP? Often no for main images—check current seller help.

Before Black Friday, shoot hero SKUs on the same backdrop roll, crop with consistent margins, and archive PNG masters by SKU code. Convert to JPG only for channels that require it. When colour variants launch, duplicate the crop settings rather than eyeballing each time.

Measure return rate on listings with halos versus reshoots; usually reshoots win for hero products.

Document crop pixel values per SKU so contractors match your house style when you outsource shooting.

Refurbished and open-box items

Include the actual unit photographed, not a stock render. Crop tightly so scratches are honest, not hidden by removal. Marketplaces ban misleading edits; removal is for backdrop clutter, not damage.

Bundle photography

When selling kits, shoot components separately, crop each, then lay out a collage in design software. Image Crop prepares parts; it does not assemble marketing banners.

International marketplaces

Some Asian platforms insist on white backgrounds with specific margins. Measure pixels after crop with Image Resize against their PDF spec sheet.

Training sales interns

Make interns run one SKU end-to-end: shoot, crop, optional removal, Image Format Converter, compress, upload to sandbox. Debrief halos before they touch the live catalog.

Returns analytics

Tag listings by whether you removed backgrounds or shot on a sweep. Quarterly, compare return reasons. Data beats arguments about whether AI removal is “good enough” for your category.

Props and lifestyle scenes often should keep backgrounds; removal is for pack-shot slots, not storytelling hero images at the top of the product detail page.

Run Image Compress after removal when PNG transparency balloons size; flatten to JPG only if the channel forbids alpha.

Jewellery photos need reflection control more than removal; a light tent cheaply beats debugging metal halos in software every season.

Food packaging with reflective film should be shot with polarised filters; software removal rarely rescues chrome wrappers on snack bars.

Vintage resale sellers should photograph flaws before removal so buyers trust the listing; authenticity beats perfectly floating sneakers on a fake white void.

Wholesale buyers comparing your pack shots to in-hand goods should recognise the same shadows; document light height in centimetres beside crop pixels in your style guide.

Marketplace A/B tests can compare conversion on removed-background main images versus lifestyle heroes; run the experiment one category at a time so seasonal inventory does not confound results.

Store employees can shoot clearance shelf tags on grey poster board so removal and crop take under a minute per SKU during slow evening shifts without a studio.

Publish the grey board SKU number in the merchandising wiki so every branch buys the same backdrop.