Scan Documents to PDF with Phone (2026)

You need to send a PDF of your driver’s license to a rental car company. You don’t have a scanner. You’ve heard of CamScanner. You also heard CamScanner was removed from Google Play in 2019 for shipping malware. So what now?

This guide covers a quieter alternative: your browser can already scan documents. Every smartphone has a high-resolution camera, every browser supports getUserMedia, and the workflow from camera frame to multi-page PDF doesn’t require installing anything. The whole thing runs on your phone or laptop with no upload to a third party.

We’ll cover why people actually need to scan documents in 2026, how getUserMedia exposes the camera to web pages (and why this is structurally privacy-safe), the perspective-correction and multi-page workflow, lighting and positioning tips that determine final quality, and how four common scanning tools (CamScanner, Adobe Scan, Microsoft Lens, imisspdf scan-pdf) compare on privacy, features, and final output quality.

Why people scan documents to PDF

The use cases are mundane and consistent. From our analytics on the scan-pdf tool and from broader industry research:

Receipts for expense reports. Reimbursement portals and accounting tools expect PDF (or PDF + structured data). A photo of a receipt is harder to attach to an expense line item than a properly cropped, deskewed scan.

Government IDs and verification documents. Visa applications, account opening (banks, brokerages, crypto exchanges), employer verification, KYC processes. These need a clean, legible PDF and they often have file size limits that a raw phone photo blows through.

Contracts and signed documents. Sign a hard-copy contract, scan it, send it back. Email a paper-signed NDA to legal. Archive the signed-by-counterparty version in your contract management system.

Whiteboard captures. A team session that produced a whiteboard full of notes. The “share to the team channel” version needs to be readable, which usually means a perspective-corrected scan, not a phone snap from an awkward angle.

Handwritten notes. Class notes, meeting notes, research notes. People scan them so they’re searchable (after OCR) and so they live alongside digital notes in Notion or Obsidian or Evernote.

Insurance and medical forms. Forms come on paper, need to be returned digitally. Filled-out paper forms scanned to PDF are an everyday workflow for anyone dealing with old-school medical or insurance providers.

Receipts for warranty claims. Multi-month gap between purchase and claim — the original receipt is faded, the digital copy you kept is gone. Scan whatever you have and submit it.

Lecture handouts and textbook pages. Students scanning a few pages of a borrowed textbook or a printed handout that the professor didn’t share digitally.

Three things connect these. The documents are usually personal or commercial. The expected output is a PDF. And the conversion happens once and never again — there’s no reason to keep the scanned image in a third-party cloud.

How getUserMedia exposes the camera to a web page

Modern browsers have a JavaScript API for capturing live video and audio: navigator.mediaDevices.getUserMedia(). Per MDN documentation, the API:

• Requires HTTPS or localhost (the API is undefined on insecure origins). Per the spec, getUserMedia can only be used in “secure contexts”.
• Requires explicit user permission via a browser-managed dialog. The dialog identifies the origin (e.g. “imisspdf.com wants to use your camera”).
• Allows the request to specify constraints (preferred resolution, front or back camera, frame rate). The browser returns the closest match the hardware supports.
• Returns a MediaStream object that the page can attach to a <video> element or draw to a <canvas> for frame capture.
• Doesn’t expose the list of cameras or microphones until permission is granted (so a page can’t enumerate your hardware without your consent).

In practice, this means a scan-pdf web page does roughly this:

1. User clicks “start scanning”. JavaScript calls getUserMedia({ video: { facingMode: 'environment' } }).
2. The browser shows a permission dialog. The user clicks allow.
3. The browser activates the back camera and starts streaming frames to a hidden <video> element on the page.
4. The user frames the document on screen and taps capture.
5. The JavaScript draws the current frame to a <canvas> at the camera’s native resolution.
6. The canvas pixels are processed (edge detection, perspective correction, contrast adjustment) entirely in the browser.
7. The processed image is added to a multi-page PDF being assembled in browser memory.
8. When the user clicks “save PDF”, the assembled PDF is offered as a download.

The crucial detail: the camera frames live in the browser tab’s memory and nothing in the API flow requires them to leave the device. A web page can choose to upload the frames to a server — many do, especially when they want server-side image processing — but the API itself doesn’t impose that. A privacy-first scanner is one that deliberately keeps the frames local.

This is why browser-based scanning is structurally privacy-safe. The same constraint that makes WebAssembly OCR (in our OCR PDF online free guide) and in-browser PDF compression (in our Compress PDF online free without losing quality guide) viable also makes scanning viable: modern browsers can do real image processing in the tab.

The native-app privacy trade-off

According to ScanLens’s coverage of on-device vs cloud OCR and Buildwize’s CamScanner alternatives analysis, the major mobile scanning apps each have a distinct privacy posture:

CamScanner. Operated by INTSIG Information Co., based in Shanghai. The app was removed from Google Play in August 2019 after Kaspersky researchers found a malicious advertising SDK shipping a trojan dropper inside the app. The malware was fixed and the app returned to the store, but the incident sits in the app’s public record. Scans uploaded to CamScanner’s cloud by default for cross-device sync and premium features. Servers are in China; users in jurisdictions with strict data-residency rules (EU, certain US sectors) should be aware that documents uploaded to CamScanner cross national borders.

Adobe Scan. Operated by Adobe. No documented breach or malware history. Scans upload to Adobe Document Cloud by default for OCR and sync with Acrobat. You can disable cloud sync but the default is upload. Adobe’s data handling is compliant with GDPR/CCPA, and they have a substantial enterprise customer base that depends on this compliance. For non-sensitive documents in an Adobe-using organization, Adobe Scan is reasonable. For documents you don’t want any third party to touch, the cloud-upload default is the issue.

Microsoft Lens. Microsoft’s mobile scanning app. Generally good privacy posture — cloud sync via OneDrive is opt-in, not default. Microsoft Lens supports saving locally without any cloud involvement. The catch: Textora and other sources have reported that Microsoft Lens has been progressively deprecated in favor of Microsoft 365 Copilot’s scan features, which are tied to a Microsoft 365 subscription. Standalone Lens still works as of 2026 but its long-term future is unclear.

imisspdf scan-pdf (and similar browser-based scanners). No app to install, no cloud sync by design — the entire workflow is local because there’s no server in the architecture. The cost is that you don’t get the convenience features that cloud-syncing apps provide (cross-device sync, organized library, history search) — once you save the PDF, the next time you scan, you start fresh.

For one-off or occasional scans of sensitive documents, the browser approach is structurally safer. For someone who scans every receipt for an entire business and wants a searchable archive synced across phone and laptop, a cloud-syncing app with explicit privacy controls is the more practical answer.

The multi-page workflow

A single-page scan is straightforward. The interesting workflow is the multi-page scan — a 12-page report, a stapled contract, a stack of receipts.

A good multi-page scanner handles four things well:

1. Sequential capture. Each page is captured, processed, and added to a working PDF in memory. The user shouldn’t have to confirm or reposition the camera between pages — just point, capture, flip page, repeat.

2. Auto-detection and capture. The best scanners detect a document on the camera frame in real time (showing a colored outline around the detected edges) and either auto-capture when the alignment is good or wait for a tap. Manual capture is more reliable in tricky lighting; auto-capture is faster when conditions are good.

3. Per-page review. After all pages are captured, the user reviews the sequence — reorder if a page is out of sequence, retake a page that came out blurry, delete a page captured by accident. Forcing the user to start over on a 12-page scan because page 9 was blurry is a tool failure.

4. Final processing. Apply consistent contrast/color processing across all pages so a 12-page scan doesn’t look like 12 different lighting conditions. Optional: convert to grayscale or black-and-white for receipt/contract use cases where color is irrelevant and the smaller file size matters. Optional: run OCR across all pages to make the final PDF searchable.

The imisspdf scan-pdf tool implements all four in-browser. The capture loop runs at the camera’s native frame rate (30 fps on most phones), edge detection runs at 5-10 fps on the live preview, and each captured frame is processed and stored in memory before the next capture. A 12-page scan typically takes 60-90 seconds end-to-end including review.

Lighting and positioning tips

Most “bad scan” outcomes have nothing to do with the scanning tool. They come from input that no tool can fully fix.

Lighting:

• Use even, diffuse light. North-facing window during daylight is ideal.
• Avoid direct sunlight (creates harsh shadows and color burn on the bright side).
• Avoid mixing light sources (window + desk lamp = color cast).
• If you’re stuck with bad room lighting, the phone’s flash is sometimes better than nothing — but it creates a hot spot in the center of the page. Use only as a last resort.

Positioning:

• Place the document flat on a contrasting surface (light document on dark desk, dark document on light desk). Edge detection works by finding the contrast boundary.
• Hold the phone parallel to the document, not tilted. Perspective correction can fix moderate tilt but not extreme angles.
• Distance: close enough that the document fills 80%+ of the frame, far enough that the camera doesn’t lose focus on the edges.
• Keep the phone steady. If you’re shaky-handed, brace your elbows on a desk or use a tripod (any tripod with a phone clamp works).

Document prep:

• Flatten folds and curls. A curved page produces a curved scan that perspective correction can’t fully fix.
• Remove staples and binding if you can. A spread-flat sheet scans better than a constrained one.
• Cover one side of double-sided documents if you can see show-through. A dark surface behind the page reduces it.

Settings:

• Use the rear camera, not the front. Higher resolution and better optics.
• Don’t zoom optically (digital zoom degrades quality). If you need closer framing, physically move the phone closer.
• For high-text-density documents, capture at the highest resolution your phone supports. For glossy receipts or simple receipts, medium resolution is fine.

A well-lit, well-positioned scan often needs no processing beyond perspective correction. A poorly-lit, poorly-positioned scan can’t be saved by even the best processing — you’ll get a usable result faster by re-capturing in better conditions than by fighting with edits.

The four tools, compared

We evaluated CamScanner, Adobe Scan, Microsoft Lens, and imisspdf scan-pdf on a realistic multi-page scan (5-page paper contract, normal office lighting, scanned at desktop with a recent iPhone). Scoring is qualitative — these are all functional tools and the choice is rarely about output quality.

Dimension	CamScanner	Adobe Scan	Microsoft Lens	imisspdf scan-pdf
Platform	iOS, Android	iOS, Android	iOS, Android	Any browser (mobile + desktop)
Install required	Yes	Yes	Yes	No
Account required	Yes (forced sign-in)	Yes (Adobe ID)	No	No
Cloud upload by default	Yes	Yes	No (opt-in to OneDrive)	No (no upload at all)
Server location	China	US/EU (Adobe Document Cloud)	Microsoft Azure (region varies)	None
Edge detection	Excellent (live overlay)	Excellent (live overlay)	Good (live overlay)	Good (live overlay, browser-based)
Perspective correction	Auto + manual	Auto + manual	Auto + manual	Auto + manual
Multi-page PDF	Yes	Yes	Yes	Yes
OCR	Yes (premium)	Yes	Yes	Yes (in-browser, free)
Watermark on free tier	Yes (free plan adds watermark)	No	No	No
File size on output	Optimized	Optimized	Optimized	Optimized
Privacy track record	Malware history (2019)	Clean	Clean	No data to compromise
Free tier limitations	Watermark, scan limits, ads	None (need Adobe ID)	None	None
Paid tier	$5-10/month	Included with Acrobat	Microsoft 365 (broader bundle)	Free

The output quality is roughly equivalent across all four for typical documents. The differences are around the experience: CamScanner is the most polished UI but the worst privacy track record; Adobe Scan integrates beautifully with the Adobe ecosystem; Microsoft Lens integrates with Microsoft 365 (and may be sunsetting); imisspdf scan-pdf has zero install/signup friction and zero upload, but you give up the cross-device library that the native apps provide.

For one-off sensitive scans, the browser tool is the structurally safer answer. For a daily-scanning workflow where library management matters, a native app is more practical — and Microsoft Lens, with its privacy-conscious defaults, is the choice we’d recommend among the native options.

Step-by-step: scanning with imisspdf scan-pdf

The workflow if you want to try it. Total time: about 90 seconds for a 5-page document on a phone, plus a couple of seconds for the page to load.

1. Open the tool. Go to imisspdf scan PDF in your phone’s browser (Safari on iPhone, Chrome on Android — any modern browser works). The page loads. The scan workflow is already there.

2. Grant camera access. Click “start scanning”. Your browser shows a permission dialog: “imisspdf.com wants to use your camera”. Click allow. The permission is per-origin and revocable from your browser’s site settings; you can revoke it any time and the tool stops working.

3. Frame the first page. Position the phone over the document. A colored outline appears on the live preview when the tool detects the document edges. Tap capture (or wait for auto-capture if it’s enabled).

4. Confirm. The captured page appears with edge points you can drag if the auto-detection got them wrong. Adjust if needed, click confirm.

5. Repeat for additional pages. Flip to page 2, capture. Flip to page 3, capture. The tool tracks your sequence — pages are stored in memory as you go.

6. Review. When you’re done, switch to review mode. You see all pages in order. Reorder by drag-and-drop, retake any page that came out badly, delete pages captured by accident.

7. Optional: OCR. If you want the result to be searchable, tap “run OCR” and pick the language. This adds a text layer over the scanned images using the same Tesseract.js engine described in our OCR guide.

8. Save. Tap “save PDF”. The assembled multi-page PDF is offered as a download. On a phone, this saves to your default downloads location (Files app on iOS, Downloads folder on Android). Share or email from there.

9. Leave the page. Closing the tab clears the in-memory pages. The PDF on your device’s storage is yours — nothing on any server.

When to use the browser tool vs a native app

Honest framing: there’s no universal winner. The right pick depends on what you’re scanning and how often.

Use a browser-based scanner (imisspdf scan-pdf or similar) when:

• The document is sensitive — IDs, signed contracts, medical forms, anything you wouldn’t email to a stranger
• You scan a few documents per month, not per day
• You’re on a phone or laptop that doesn’t have your usual scanning app installed
• You don’t want to install another app for an occasional task
• You don’t want to create another account or sign in
• You’re in a jurisdiction with data-residency requirements that cloud-syncing apps can’t meet

Use Microsoft Lens when:

• You already use Microsoft 365 and want OneDrive integration
• You want a free, native app with privacy-conscious defaults (no forced cloud sync)
• You scan frequently enough that the library management is valuable

Use Adobe Scan when:

• You already use Adobe Acrobat and the integration matters
• You want best-in-class auto-cropping and processing in a polished native app
• You’re scanning non-sensitive material and the cloud upload default isn’t an issue

Avoid CamScanner if:

• The document is sensitive
• You’re in a jurisdiction with strict data-residency or import-control rules
• You’re an organization with vendor risk policies that have flagged the 2019 incident

Try scanning in your browser

If you want to scan a few documents without installing anything or uploading anything, try imisspdf scan PDF → on your phone or laptop. No signup, no install, no cloud upload, no watermark. Works on iOS Safari, Android Chrome, desktop browsers — anywhere your browser can access a camera.

For frequent business scanning where library management matters, Microsoft Lens remains the most privacy-conscious of the native apps and is genuinely well-built. The frame that works best across all of these: decide per document. Confidential paperwork: browser scan, no upload. Day-to-day receipt logging for an expense system: whichever native app integrates with your accounting tool.

Frequently asked questions

The FAQ block at the top of this article covers the most common questions about phone-camera document scanning. For related coverage, see How to OCR a scanned PDF and Compress PDF online free without losing quality — the natural next steps after capturing a scan. Healthcare and legal professionals often scan sensitive documents on mobile — see our PDF Tools for Healthcare and PDF Tools for Lawyers guides for privacy-first workflows. For comparison with CamScanner/Adobe Scan, our iLovePDF safety review covers the upload-vs-local trade-off in more depth.

Sources

• MDN: MediaDevices.getUserMedia()
• web.dev: Capture audio and video in HTML5
• ScanLens: On-Device vs Cloud OCR — Privacy, Speed, Accuracy
• Buildwize: Best CamScanner Alternatives in 2026 — Safer & Better Options
• Textora: Microsoft Lens Is Gone — Best Replacements in 2026
• Software Advice: CamScanner vs Microsoft Lens — 2026 Comparison
• Adobe Scan privacy and data handling documentation
• Kaspersky: CamScanner malware analysis (2019)

Frequently asked questions

Related articles