Orris

Change background to white

I don't see any file, canvas, or UI context attached to your message. Could you clarify what you'd like to change the background of?
Asking for Preferences

What would you like to change the background to white on?

I don't have enough context to proceed. To change a background to white, I need to know:
  1. What file or project - please share the file (HTML, PPTX, image, etc.) or paste the relevant code.
  2. Which element - the whole page/slide, a specific section, or a specific element?
Please share the file or code you're working with and I'll make the change right away.

Change background to white

Let me read the PDF skill first, then process your file.
Reading File
Reading File
Now let me look at the PDF to understand what we're working with.
Reading File
It's a scanned/image-based PDF. I'll use pikepdf to add a white background rectangle behind all content on each page.
Running Command

python3 - << 'EOF' import subprocess result = subprocess.run(['python3', '-c', 'import pikepdf; print("pikepdf ok"); import pypdf; print("pypdf ok"); import fitz; print("fitz ok")'], capture_output=True, text=True) print(result.stdout) print(result.stderr) EOF

Running Command

python3 -c "import fitz; print('fitz/pymupdf ok, version:', fitz.version)"

Running Command

python3 -c "import pypdf; print('pypdf ok')" && python3 -c "from PIL import Image; print('PIL ok')" && which gs && gs --version

Running Command

python3 -c " import pypdf reader = pypdf.PdfReader('/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47.pdf') print('Pages:', len(reader.pages)) page = reader.pages[0] print('MediaBox:', page.mediabox) print('Resources keys:', list(page.get('/Resources', {}).keys()) if page.get('/Resources') else 'None') # Check if there are images (XObjects) res = page.get('/Resources') if res: xobj = res.get('/XObject') if xobj: print('XObjects:', list(xobj.keys())) "

It's a 5-page PDF with images (scanned document). I'll use PIL to convert each page's image to white background (replacing dark/gray background) and rebuild the PDF.
Running Command

python3 << 'EOF' import pypdf import io from PIL import Image src = '/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47.pdf' reader = pypdf.PdfReader(src) for i, page in enumerate(reader.pages): res = page.get('/Resources') if res: xobj = res.get('/XObject') if xobj: for key in xobj.keys(): obj = xobj[key].get_object() subtype = obj.get('/Subtype') print(f"Page {i+1}, {key}: Subtype={subtype}, Filter={obj.get('/Filter')}, Width={obj.get('/Width')}, Height={obj.get('/Height')}, CS={obj.get('/ColorSpace')}") EOF

The PDF has JPEG images - X2 is the main scanned page image and X1 is likely a small header/logo. I'll process each page image: whiten the background (replace the non-white/gray scanner background with pure white) and rebuild the PDF.
Running Command

python3 << 'EOF' import pypdf import io from PIL import Image import numpy as np src = '/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47.pdf' reader = pypdf.PdfReader(src) # Extract the main image from page 1 to preview what it looks like page = reader.pages[0] res = page.get('/Resources') xobj = res.get('/XObject') obj = xobj['/X2'].get_object() data = obj.get_data() img = Image.open(io.BytesIO(data)) arr = np.array(img) print("Image shape:", arr.shape) print("Min pixel:", arr.min(), "Max pixel:", arr.max()) print("Mean pixel:", arr.mean()) # Sample corners to see background color print("Top-left 5x5 mean:", arr[:5, :5].mean()) print("Top-right 5x5 mean:", arr[:5, -5:].mean()) print("Bottom-left 5x5 mean:", arr[-5:, :5].mean()) EOF

Running Command

python3 << 'EOF' # Let's look at the pixel value distribution to understand the background import pypdf import io from PIL import Image import numpy as np src = '/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47.pdf' reader = pypdf.PdfReader(src) page = reader.pages[0] res = page.get('/Resources') xobj = res.get('/XObject') obj = xobj['/X2'].get_object() data = obj.get_data() img = Image.open(io.BytesIO(data)) arr = np.array(img) gray = arr.mean(axis=2) # Histogram of brightness bins = [0,50,100,150,180,200,220,240,255] hist, edges = np.histogram(gray, bins=bins) for i in range(len(hist)): print(f"{edges[i]:.0f}-{edges[i+1]:.0f}: {hist[i]} pixels ({hist[i]/gray.size*100:.1f}%)") EOF

The background is a mix of grays (typical scanner output). I'll use a threshold approach: pixels brighter than ~150 (background/paper) will become pure white, while darker pixels (text/ink) will be preserved. This cleans up scanner gray to white.
Running Command

python3 << 'PYEOF' import pypdf from pypdf import PdfWriter, PdfReader from pypdf.generic import ( ArrayObject, DecodedStreamObject, DictionaryObject, NameObject, NumberObject, RectangleObject ) import io from PIL import Image import numpy as np src = '/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47.pdf' out_path = '/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47 white bg.pdf' reader = PdfReader(src) writer = PdfWriter() def whiten_image(jpeg_bytes, threshold=150): """Replace pixels brighter than threshold with pure white.""" img = Image.open(io.BytesIO(jpeg_bytes)).convert('RGB') arr = np.array(img, dtype=np.uint8) # Compute grayscale brightness gray = arr.mean(axis=2) # Mask: background pixels (bright gray) -> white mask = gray > threshold arr[mask] = [255, 255, 255] # Save back to JPEG bytes out = io.BytesIO() Image.fromarray(arr).save(out, format='JPEG', quality=92) return out.getvalue() for page_num, page in enumerate(reader.pages): print(f"Processing page {page_num+1}...") res = page.get('/Resources') if res: xobj = res.get('/XObject') if xobj: for key in list(xobj.keys()): obj = xobj[key].get_object() subtype = obj.get('/Subtype') if subtype == '/Image': width = int(obj.get('/Width')) height = int(obj.get('/Height')) # Only process the main large page image (X2), skip small header (X1) if width > 500 and height > 500: jpeg_data = obj.get_data() new_jpeg = whiten_image(jpeg_data) # Replace stream data obj._data = new_jpeg obj[NameObject('/Length')] = NumberObject(len(new_jpeg)) print(f" Whitened {key}: {width}x{height}, {len(jpeg_data)} -> {len(new_jpeg)} bytes") writer.add_page(page) with open(out_path, 'wb') as f: writer.write(f) print(f"\nDone! Saved to: {out_path}") PYEOF

All 5 pages processed. The gray scanner background has been replaced with pure white while preserving the text and content.
[New Doc 05-26-2026 16.47 white bg.pdf](/home/daytona/workspace/attachments/c8268bae-aeff-410a-ac9c-3e807354df20/New Doc 05-26-2026 16.47 white bg.pdf)
All 5 pages have been processed - pixels brighter than the threshold (the gray scanner background) were replaced with pure white, while darker pixels (text, ink) were kept intact.
This is a shared conversation. Sign in to Orris to start your own chat.