Here’s how to get the best results when scanning your bookshelves.
Taking good photos
Lighting
Good lighting is the most important factor for accurate OCR.
| Condition | Quality | Tip |
|---|---|---|
| Natural daylight | Excellent | Best results |
| Bright room light | Good | Avoid shadows |
| Dim lighting | Poor | Use flash or add light |
| Direct sunlight | Variable | Watch for glare |
Angle
- Straight-on works best
- Slight angle (up to 15°) is acceptable
- Avoid steep angles that distort text
Distance
- Close enough that spine text is readable
- Far enough to capture 5-15 books per photo
- Typical distance: 30-60 cm (1-2 feet)
Focus
- Ensure text is sharp, not blurry
- Tap to focus on your phone’s camera
- Hold steady while capturing
Scanning modes
Spine mode (default)
Best for: Bookshelves with visible spines
- Take photo of bookshelf
- App detects individual spine regions
- OCR extracts text from each spine
- Text is searched against book databases
Tips:
- Works best with clearly printed spines
- Handles both horizontal and vertical text
- Multiple books per photo
Barcode mode
Best for: Books with visible ISBN barcodes
- Enable “Barcode Mode” in settings
- Point camera at book barcode
- ISBN is detected automatically
- Direct lookup in book databases
Tips:
- Most accurate method
- Works one book at a time
- Barcode must be visible and clear
Troubleshooting scans
”No books detected”
Causes:
- Photo too dark
- Spines too small in image
- Poor contrast
Solutions:
- Add more light
- Get closer to bookshelf
- Ensure spines face camera
”Poor OCR results”
Causes:
- Blurry photo
- Small or decorative fonts
- Faded or damaged spines
Solutions:
- Hold camera steady
- Use Google Cloud Vision (more accurate)
- Manually correct results
”Wrong book matches”
Causes:
- Common words on spine
- Generic titles
- Multiple editions
Solutions:
- Use barcode mode for accuracy
- Edit and re-search
- Verify against your actual books
”Barcode not detected”
Causes:
- Barcode damaged or dirty
- Glare on barcode
- Camera too far
Solutions:
- Clean the barcode
- Reduce glare
- Move closer
Optimal shelf setup
For best results:
- Arrange books vertically with spines facing out
- Remove dust jackets if they cover spine text
- Ensure even lighting across the shelf
- Group by size if possible (easier detection)
Batch scanning
For large libraries:
- Scan one shelf at a time
- Review each batch before exporting
- Merge exports in Libiry
This approach:
- Keeps photo quality high
- Makes review manageable
- Reduces memory usage
Mobile tips
Android
- Use Chrome for best PWA support
- Install app for faster access
- Grant camera permission permanently
iOS
- Use Safari (required for PWA)
- Add to Home Screen for app-like experience
- Photos may need good lighting (no flash)
Phone camera vs. desktop upload — why results may differ
When you scan a bookshelf with your phone camera, the results may be different from uploading the same shelf as a photo on desktop. This is normal and has several causes:
- Resolution: Phone cameras compress the photo before sending it to the app. The uploaded photo on desktop is usually the original, uncompressed file with more detail.
- Focus and sharpness: A phone camera may not focus perfectly on every spine, especially at the edges of the frame. A carefully taken photo that you upload later may be sharper.
- Lighting: Phone photos taken in indoor lighting often have more noise and less contrast than well-lit photos.
- Camera angle: A slight angle or tilt causes perspective distortion, making some spines harder to read.
Tips for better phone results:
- Hold the phone steady and straight in front of the bookshelf
- Make sure there is good, even lighting (avoid shadows across the spines)
- Take the photo from a distance where all spines are readable to the human eye
- If results are poor, try taking multiple photos of smaller sections of the shelf
- Use Google Cloud Vision instead of Tesseract for significantly better results on phone photos