Removing the background from a photo is easy in theory. In practice, different types of photos present different challenges. A headshot with flyaway hair requires a different approach than a white sneaker on a white table.
This guide covers the specific techniques and best practices for the two most common photo background removal tasks: portraits and product photography. You'll learn how to get clean results on the first try and handle the tricky edge cases each category presents.
Portrait Background Removal
Portraits are the most challenging subject for background removal because of one thing: hair. The boundary between a person's hair and the background contains thousands of fine, semi-transparent strands that occupy a single-pixel-wide zone between "subject" and "background."
The Hair Challenge
Older background removal tools handle hair poorly — they either clip the strands too aggressively (creating a harsh, unnatural outline) or include too much background (leaving a visible color fringe). This is the main reason people spend 15+ minutes with Photoshop's Refine Edge tool.
BestPNG's photo background remover uses BiRefNet, which processes fine details through a dedicated high-resolution reference path. The result: individual hair strands are preserved with natural transparency, not clumped into a solid mass.
Tips for Portrait Photos
Shoot against a contrasting background. If the person has dark hair, use a light background. Blonde or gray hair works better against a dark background. This gives the AI maximum information to distinguish hair from background.
Ensure even lighting on the subject. Harsh shadows across the face or body can confuse edge detection. Soft, diffused lighting produces the cleanest cutouts.
Avoid backgrounds that match clothing. A person wearing a green jacket against green foliage is harder to process than the same jacket against a gray wall. Color similarity between subject and background is the main source of errors.
Include the full head in frame. Cropped hair at the top of the frame forces the AI to guess where the boundary is. Keep some margin above the head.
Watch for color spill. When a person is photographed near a strongly colored wall, reflected light can tint their skin and hair. This color contamination can remain visible after background removal. Shooting further from colored surfaces helps.
Common Portrait Use Cases
- Professional headshots — LinkedIn, company websites, team pages
- ID and passport photos — Solid white or blue background requirements
- Social media — Profile pictures, story cutouts, thumbnail overlays
- Event photography — Isolating individuals from group shots or busy venues
Product Photography Background Removal
Products present different challenges than portraits. Edges are typically sharp and well-defined, but shadows, reflections, and similar colors between the product and background can cause issues.
Shadow Handling
Product shadows add depth and realism. The question is: should you keep them or remove them?
Remove shadows when: You need a pure transparent background for marketplace listings (Amazon, eBay) or when the product will be placed on a new background with its own lighting.
Keep shadows when: The product will be placed on white and you want a natural drop shadow. In this case, remove the background first with BestPNG, then add a consistent drop shadow in your design tool.
Most AI background removers, including BestPNG, remove the background and natural shadows together. For a clean, flexible workflow, this is actually ideal — you can always add a controlled shadow later, but removing an inconsistent natural shadow is much harder.
Reflection Handling
Products with glossy or reflective surfaces (watches, electronics, jewelry) sometimes show reflections of the original background. After removing the background, these reflections can look odd because they contain colors from a surface that no longer exists.
For most use cases, the reflection is minor enough to ignore. For high-end product photography, reshoot the product on a neutral surface or use a photo editor to paint over the reflection area.
Tips for Product Photos
Use a contrasting background for the shoot. A white product on white paper is the worst case for any background remover. Shoot on a gray, blue, or contrasting solid surface — the AI will produce a cleaner edge, and you can place the cutout on white afterward.
Light products evenly. Soft, even lighting minimizes harsh shadows and ensures the product edges are clearly defined. A lightbox or two softbox lights are well-suited for small products.
Photograph at high resolution. More pixels mean sharper edges in the cutout. For e-commerce, most platforms recommend at least 1000x1000 pixels, and many prefer 2000x2000 or higher.
Keep the product stable and centered. A clean composition makes automatic detection easier. Use a tripod and fill the frame without cropping the product.
Remove stickers, tags, and dust before shooting. It's much easier to prep the physical product than to edit these out digitally.
E-commerce Platform Requirements
Each major marketplace has specific image requirements. Here's what the biggest platforms expect:
| Platform | Background | Min Size | Format | Other |
|---|---|---|---|---|
| Amazon | Pure white (#FFFFFF) | 1000x1000 | JPG or PNG | Product fills 85%+ of frame |
| eBay | White or light gray | 500x500 | JPG or PNG | No borders or watermarks |
| Shopify | Consistent across listings | 2048x2048 recommended | JPG or PNG | Square images preferred |
| Etsy | White or natural | 2000x2000 recommended | JPG or PNG | First image on white background |
| AliExpress | White | 800x800 | JPG or PNG | No text overlays on main image |
The workflow: shoot the product, remove the background with BestPNG to get a transparent PNG, then place it on the required white background at the correct dimensions. If you need JPG output, convert the PNG to JPG as the final step.
Batch Processing Workflow
If you have 20, 50, or 200 product photos to process, doing them one by one is impractical. Here's an efficient batch workflow:
-
Prepare your images. Ensure consistent naming, resolution, and format. Rename files sequentially (product-001.jpg, product-002.jpg, etc.) for easy tracking.
-
Batch upload to BestPNG. Use BestPNG's batch processing feature to upload and process multiple images at once. The AI handles each image independently, so mixed subjects (different products, different backgrounds) work fine.
-
Review results. Scan through the outputs. Most will be clean on the first pass. Flag any that need manual attention.
-
Post-process if needed. For images that need touch-up, open in Photoshop or GIMP for quick edge refinement. For images destined for marketplaces, place on white and resize to platform specifications.
-
Compress for web. Product images on e-commerce sites need to load fast. Run the final images through a PNG compressor or convert to optimized JPG to reduce file sizes.
This workflow can process 100+ product images in under an hour, compared to 2-3 days of manual work.
Advanced: White Products on White Backgrounds
This is the hardest scenario for any background remover. When the product and background share the same color, the AI has minimal contrast to work with.
Solutions:
- Shoot on a colored surface first. Gray paper is ideal. Remove the background, then place the cutout on white.
- Add subtle edge lighting. A thin rim light along the product edges creates separation even when the colors match.
- Increase the background distance. Moving the product further from the background creates a natural shadow that helps the AI identify the boundary.
Frequently Asked Questions
How do I remove the background from a group photo to isolate one person? Upload the group photo to BestPNG. The AI will attempt to isolate all foreground subjects. For isolating a specific individual, crop the image to focus on that person before uploading, or use Photoshop to select and extract them manually.
Can I remove the background from a product photo taken with my phone? Absolutely. Modern phone cameras produce images at 12 MP or higher, which is more than enough resolution. Ensure good lighting and a contrasting background for best results.
What about images with multiple products? BestPNG will keep all foreground objects and remove the background behind them. If you need individual product cutouts, photograph each product separately or crop the image before processing.
How do I handle products with holes or negative space (like a chair or a mug handle)? BiRefNet handles interior holes and negative space well. The AI recognizes that the background visible through a mug handle or between chair legs should also be removed. Complex topology like this is one area where modern AI models significantly outperform older tools.
What if the product edges look slightly rough after removal? For hard-edged products like electronics or boxes, the edges should be crisp. If they appear rough, the source image may have been slightly out of focus. Try re-shooting with a tripod and sharper focus, or use a PNG sharpening tool on the result.
Conclusion
Portrait and product photos each have their own background removal challenges, but modern AI handles both remarkably well. For portraits, the key is hair edge quality — and BiRefNet excels here. For products, the key is clean, consistent edges and proper shadow handling.
BestPNG's photo background remover handles both categories effectively. Upload a portrait, get natural hair edges. Upload a product photo, get a clean cutout ready for any marketplace or design project.