MakeACopy
de.schliweb.makeacopy
Zapstore _@zapstore.dev Republished from GitHub / F-Droid by the Zapstore main account.
MakeACopy is an open-source document scanner app for Android that allows you to digitize paper documents with OCR functionality. The app is designed to be privacy-friendly, working completely offline without any cloud connection or tracking.
First release: Sep 24, 2025, 45 total releases.
Most recent release: Jul 9, 2026.
Appears in 1 app stack.
0 sats / 0 zaps received in the past year.
Sats Received
Underlying data available via MCP: app_zaps, app_releases.
Zap Count
Underlying data available via MCP: app_zaps, app_releases.
Releases
- Jul 9, 2026 4.6.1MakeACopy 4.6.1 * More accurate corner detection in the crop view: corners now stay on the document edges after layout changes. * New edge-snap corner refinement for sharper document boundaries. * Improved black & white conversion: cleaner pages, fewer speckles, better readability of faint text and low-res scans.
- Jun 19, 2026 4.6.0TEST-BUILD MakeACopy 4.6.0 * Full-text search across your scan library, now including recognized OCR text (Room/FTS5). * Search results are highlighted in the scan list for faster navigation. * OCR text is stored and indexed during export so documents stay searchable. * New experimental PaddleOCR v6 Small recognition model (opt-in) with language-aware routing. * Various stability and performance improvements under the hood.
- Jun 19, 2026 4.5.1MakeACopy 4.5.1 * Fixed a device-dependent aspect-ratio mismatch: preview, live overlay and the captured image now share the same field of view. * Live document corners align more accurately with the camera preview. * More stable crop screen: corner detection no longer runs twice during layout, keeping the accurate result. * Consistent thumbnails across lists and export (no more cropped previews). * Unified text sizes for a cleaner, more polished UI.
- Jun 14, 2026 4.5.0MakeACopy 4.5.0 * Material You dynamic colors (Android 12+) and Material 3 dialogs. * Themed (monochrome) launcher icon. * Workflow stepper across Scan, Crop, OCR and Export. * Search in the scan library. * New "Share text" action and export format selector (PDF/JPEG). * Option to skip edge detection, smoother transitions and haptic feedback. * Fixed duplicate dialogs on rapid taps and redundant edge re-detection.
- May 24, 2026 4.2.0MakeACopy 4.2.0 * Redesigned export: Color, Grayscale and B&W for JPEG/PDF. * New cleanup options for crisper scans. * PaddleOCR: optional best-quality mode for better text detection. * More robust PDF creation and OCR tests.
- May 22, 2026 4.1.3MakeACopy 4.1.3 * Release metadata now distinguishes the F-Droid standard build from the Paddle OCR build more clearly. * The Paddle build keeps the plain 4.1.3 version name to match F-Droid metadata requirements.
- May 17, 2026 4.1.1-paddleMakeACopy 4.1.1 * Published Store listings now use the PaddleOCR flavor; Tesseract standard builds remain on GitHub. * Fixed crop handling so OpenCV fallback rectangles no longer override better document-corner estimates.
- May 8, 2026 3.8.0MakeACopy 3.8.0 * Crop: aspect ratios (A3/A4/A5, Letter, Legal, custom), snap-to-right-angle, better edge drag, zoom/pan * Camera: sharper shots via AF/AE/AWB pre-lock; faster preview via NV21→RGBA fast path + pooling * Export: re-edit crop from preview; new high-pass "clean" modes (gray/color); zoomable preview * OCR: 2 modes + smarter policies * Offline only
- May 1, 2026 3.7.2MakeACopy 3.7.2 * Improved automatic aspect ratio of perspective-corrected scans: tilted A4 letters, long receipts and packaging (e.g. Tetra Pak) come out far less squashed * Uses a projective aspect-ratio estimation (Zhang & He, 2006) with graceful fallback to the previous heuristic for near-fronto-parallel captures * Fully offline, no cloud or tracking
- Mar 8, 2026 3.5.0MakeACopy 3.5.0 * New: Inbox Mode – configure a default export folder for frictionless scanning, no file picker needed (ideal for paperless-ngx, Nextcloud, Syncthing workflows). Guide: https://github.com/egdels/makeacopy/blob/main/docs/inbox_mode_guide_en.md * Fixed: File picker now remembers the last used folder correctly (improved initial URI handling for better OEM compatibility) * Fixed: Crash (OutOfMemoryError) during JPEG-compressed PDF export caused by alpha channel extraction in large images * Fully offline, no cloud or tracking
- Mar 6, 2026 3.4.0MakeACopy 3.4.0 (Release) * Added Turkish language support for OCR * Added configurable cleanup for completed scans (by age, count, or storage) * Independent storage locations for import and export * Fully offline, no cloud or tracking Hint: This is a release candidate - feedback and bug reports are welcome!
- Mar 3, 2026 3.3.1MakeACopy 3.3.1 (Release) * Fixed crash on certain Android 15 devices (e.g. Redmi Note 13 Pro 5G) when starting OCR processing — switched from OpenMP-enabled Tesseract to non-OpenMP variant to avoid sched_setaffinity crashes * Fully offline, no cloud or tracking
- Feb 25, 2026 3.2.1MakeACopy 3.2.1 * Improved shadow removal for scanned documents (broader illumination correction) * Enhanced local contrast (CLAHE) for sharper, more readable text * Switched to Gaussian-weighted adaptive thresholding for better document binarization * Improved binarization quality for low-resolution images (e.g. from autoscan) * PDF Classic B/W mode now includes shadow removal and contrast enhancement * Faster image capture by removing unnecessary post-focus delay * Fully offline, no cloud or tracking
- Feb 14, 2026 3.1.0MakeACopy 3.1.0 (Release-Candidate) * Updated OpenCV to 4.13.0 for better image processing * Updated ONNX Runtime to v1.24.1 for faster edge detection * Improved code quality and maintainability * Performance optimizations throughout the app * Fully offline, no cloud or tracking Hint: This is a release candidate. Please report any issues you find.
- Feb 12, 2026 3.0.1MakeACopy 3.0.1 (Bug-Fix) * Fixed repeated ML model loading during live camera analysis. * Document detection now starts significantly faster. * Fully offline, no cloud or tracking.
- Feb 9, 2026 3.0.0MakeACopy 3.0.0 (Release) * Replaced DocAligner with a self-trained ONNX corner detection model. * Model size reduced from ~12,6 MB to ~4,8 MB. * Improved robustness and accuracy of document detection. * Fully offline, no cloud or tracking.
- MakeACopy 2.6.0 (Release) * Switched OCR engine to tesseract4android-openmp for better performance. * Fixed OCR edge case where high confidence was reported with no detected text. * Empty OCR results are now handled deterministically and normalized to confidence = 0. * Removed misleading OCR score toasts when no text is detected. * Added a clear “No text detected” message. * Improved OCR retry and rotation handling for empty results. * Stability and correctness improvements in the OCR pipeline.
- MakeACopy 2.5.1 Bug fixes - Fixed an issue where importing an image (gallery) or PDF could sometimes show an old OCR result. - Update Tesseract dependency group and URL
- MakeACopy 2.5.0 (Release) What’s new - Improved document edge detection and perspective correction, especially for distorted or angled pages. - Faster image capture: reduced focus timeout and paused live analysis during capture for improved responsiveness. - Enhanced corner detection with toggle support for OpenCV/ONNX detectors, Hough line-based fallback, dynamic threshold adjustment, low-light handling,
More…
and consensus-based fusion for improved precision. Bug fixes - Fixed minimap thumbnail not reflecting user-defined image rotation in OCR review. - Fixed connected components analysis for OCR preprocessing (was incorrectly analyzing background instead of text). - Fixed memory leak in border noise clearing (SubMat resources now properly released). - Fixed inconsistent angle validation in ONNX corner detection. - Improved resource management for OpenCV contour processing. - Fixed HTML entity encoding in OCR text display (e.g., ' now correctly shows as apostrophe in OCR Review and PDF export). - MakeACopy 2.4.1 (Release) - Multi-language OCR: Select up to 2 languages simultaneously for documents with mixed scripts (e.g., Persian + English, Arabic + French, German + English). - Combined dictionaries: OCR post-processing now uses combined dictionaries from all selected languages for improved error correction in multilingual documents. - Improved multi-language support: Enhanced handling of RTL + LTR script combinations and proper whitelist management for mixed-script recognition. - OCR Review improvements: - Dictionary-based word suggestions using Levenshtein similarity with OCR-specific weighted substitutions (0↔o, 1↔l, 5↔s). - Re-OCR single words: Change the language for a word and re-run OCR recognition for that specific word. - Confidence updates: After Re-OCR, the confidence score is recalculated and displayed with color-coded highlighting. - Language persistence: Word-level language assignments are now correctly saved and restored when reopening the Review screen. - UI polish: Improved button spacing in the inline edit dialog.
- MakeACopy 2.4.0 (Release) - Arabic and Persian OCR support: Full recognition with AUTO page segmentation and disabled Latin whitelist for proper Arabic script detection. - New dictionaries: Added Arabic and Persian frequency dictionaries for improved OCR post-processing. - 21 languages now supported: Extended dictionary download to cover all supported OCR languages. - OCR quality improvements: Dynamic Best/Fast model detection, enhanced adaptive preprocessing with Wolf and NICK thresholds, and dictionary-based error correction. - PDF text layer improvements: Dynamic line tolerance based on median word height, proximity-based word clustering, and improved line sorting by average Y-position. - BiDi support: RTL detection via Unicode directionality analysis, correct right-to-left sorting for Arabic and Persian scripts. - Noto Naskh Arabic font: Added for proper Arabic and Persian script rendering in PDFs. - Fallback dictionary loading: Supports both compressed (.txt.gz) and uncompressed (.txt) dictionaries. - PDF import: Import PDF files via the image picker button. Single-page PDFs are rendered directly; multi-page PDFs show a page selection dialog with thumbnail previews.
More…
- Bug fix: Fixed crash when closing PDF resources (IllegalStateException: Document already closed). - Bug fix: Fixed OpenCV assertion error in classic B/W export mode when adaptive threshold was skipped. - Bug fix: Improved readability of B/W PDFs at low DPI settings. The despeckle algorithm now scales its aggressiveness based on target DPI, preserving fine text details at lower resolutions. - Bug fix: Fixed issue where old OCR results were displayed after switching from PDF import to camera capture when cropping was skipped. - MakeACopy 2.3.1 (Release) - Improved accessibility: new and expanded A11y guidance, including haptic feedback and clearer hints. - Reorder pages (A11y-friendly): actions to move pages left or right. - Collections / indexing: completed scans are now automatically placed in a dedicated default collection, making them easier to find. - Unified rotation handling: more consistent image rotation across the entire workflow (viewing, processing, export). - OCR auto-rotate improved and clarified: clearer option explanation and better handling of detected rotation during export.
- MakeACopy 2.2.2 (Release) This update fixes a versionCode mismatch introduced in 2.2.0/1 that prevented binary-identical reproducible builds across ABI splits and F-Droid. No functional changes were made to the scanning engine, camera, OCR, or UI. - Camera rebuilt for broad device compatibility with dynamic 3‑stage binding. Most OEM workarounds removed; targeted handling remains for Sony on Android 15+. - Self‑healing preview to recover black screens. - Adaptive torch exposure based on live luminance; paused during corner preview to avoid darkening. - Faster frame analysis from the Y‑plane with lower CPU usage. - Cleaner lifecycle and improved recovery logging; key device/capability logs at INFO.
More…
- Share logs action collects only this app’s logs and adds a compact header. - Geometry fix: rectScore() angle calculation corrected for better contour ranking. - Version info in the shared log header now derives from the manifest at runtime (PackageInfo), ensuring accurate reporting accross ABI splits and reproducible builds. - Simplified build output by disabling non-deterministic ArtProfile tasks (baseline.prof), improving reproducibility for F-Droid and third-party rebuilders. Notes - Fixed camera preview black screen on Sony Xperia (Android 15+) and Huawei. - Improved flashlight stability on Xiaomi/POCO. - No new dangerous permissions. - MakeACopy 2.2.1 (Release) - Camera rebuilt for broad device compatibility with dynamic 3‑stage binding. Most OEM workarounds removed; targeted handling remains for Sony on Android 15+. - Self‑healing preview to recover black screens. - Adaptive torch exposure based on live luminance; paused during corner preview to avoid darkening. - Faster frame analysis from the Y‑plane with lower CPU usage. - Cleaner lifecycle and improved recovery logging; key device/capability logs at INFO. - Share logs action collects only this app’s logs and adds a compact header. - Geometry fix: rectScore() angle calculation corrected for better contour ranking. - Version info in the shared log header now derives from the manifest at runtime (PackageInfo), ensuring accurate reporting accross ABI splits and reproducible builds.
More…
Notes - Fixed camera preview black screen on Sony Xperia (Android 15+) and Huawei. - Improved flashlight stability on Xiaomi/POCO. - No new dangerous permissions. - MakeACopy 2.2.0 (Release) - Camera rebuilt for broad device compatibility with dynamic 3‑stage binding. Most OEM workarounds removed; targeted handling remains for Sony on Android 15+. - Self‑healing preview to recover black screens. - Adaptive torch exposure based on live luminance; paused during corner preview to avoid darkening. - Faster frame analysis from the Y‑plane with lower CPU usage. - Cleaner lifecycle and improved recovery logging; key device/capability logs at INFO. - Share logs action collects only this app’s logs and adds a compact header. - Geometry fix: rectScore() angle calculation corrected for better contour ranking.
More…
Notes - Fixed camera preview black screen on Sony Xperia (Android 15+) and Huawei. - Improved flashlight stability on Xiaomi/POCO. - No new dangerous permissions. - MakeACopy 2.1.1 (Release) - Language list now shows which Tesseract model is active per language: “Fast” (built-in) or “Best” (imported) - New option in OCR settings to delete an imported “Best” model for the currently selected language - After importing or deleting a model, the primary action switches to “Process” so you can quickly re-run OCR - New recognition mode selector with three options: Original (no preprocessing), OpenCV Quick (fast), OpenCV Robust (high quality); your choice is saved - New “Explain recognition modes” dialog that describes the differences in simple terms - Robust mode rebuilt for top accuracy on difficult pages (deskew, illumination/shadow normalization, edge‑preserving denoise, smart upscaling) and now feeds a high‑quality grayscale image to Tesseract to preserve fine details - Settings button is disabled while OCR is processing for fewer accidental taps - Dialog button contrast improved for Dark Mode
More…
- Minor stability and UI fixes - JPEG exporter cleaned up to consistently use OpenCVUtils across all modes, reducing duplicated image-processing code. - New export option for PDF/JPEG: "OCR robust preprocessing". Uses the same robust preprocessing as OCR; wired through the full export pipeline. Notes - “Best” models are detected via file presence/size in no_backup/tessdata; built-in assets are treated as “Fast” - No internet connection is required for OCR; models are stored on-device - MakeACopy 2.1.1 (Release Candidate) - Language list now shows which Tesseract model is active per language: “Fast” (built-in) or “Best” (imported) - New option in OCR settings to delete an imported “Best” model for the currently selected language - After importing or deleting a model, the primary action switches to “Process” so you can quickly re-run OCR - New recognition mode selector with three options: Original (no preprocessing), OpenCV Quick (fast), OpenCV Robust (high quality); your choice is saved - New “Explain recognition modes” dialog that describes the differences in simple terms - Robust mode rebuilt for top accuracy on difficult pages (deskew, illumination/shadow normalization, edge‑preserving denoise, smart upscaling) and now feeds a high‑quality grayscale image to Tesseract to preserve fine details - Settings button is disabled while OCR is processing for fewer accidental taps - Dialog button contrast improved for Dark Mode
More…
- Minor stability and UI fixes Notes - “Best” models are detected via file presence/size in no_backup/tessdata; built-in assets are treated as “Fast” - No internet connection is required for OCR; models are stored on-device Note: This is an RC build; feedback is welcome to improve stability and polish. - MakeACopy 2.1.0 (Release) - Rotate in cropping screen and updated workflow: you can now rotate directly on the cropping screen. - Improved OCR for mis-rotated images: the app now tries additional 90° steps (0°, 90°, 180°, 270°) and automatically selects the best result based on mean confidence (with text length as a tiebreaker). - New early-exit threshold: if the first pass already reaches a mean confidence of >= 55, the extra rotation attempts are skipped to deliver faster results. - Correct bounding boxes: each rotation attempt maintains an appropriate transform so word boxes are mapped accurately in the UI and export. - Stability and performance: OCR runs on a single background worker per job, with proper cancellation checks and clean Tesseract shutdown to reduce resource usage. - Crop handles stay within the image: trapezoid selection corners are now hard-clamped to the visible image bounds (FIT_CENTER), preventing dragging outside the image. - Implemented chronological insertion for pages added from “Completed Scans.” When multiple items are picked, they are now sorted by their creation timestamp before being added to the current export session.
- MakeACopy 2.1.0 (Release Candidate) - Rotate in cropping screen and updated workflow: you can now rotate directly on the cropping screen. - Improved OCR for mis-rotated images: the app now tries additional 90° steps (0°, 90°, 180°, 270°) and automatically selects the best result based on mean confidence (with text length as a tiebreaker). - New early-exit threshold: if the first pass already reaches a mean confidence of >= 55, the extra rotation attempts are skipped to deliver faster results. - Correct bounding boxes: each rotation attempt maintains an appropriate transform so word boxes are mapped accurately in the UI and export. - Stability and performance: OCR runs on a single background worker per job, with proper cancellation checks and clean Tesseract shutdown to reduce resource usage. - Crop handles stay within the image: trapezoid selection corners are now hard-clamped to the visible image bounds (FIT_CENTER), preventing dragging outside the image. - Implemented chronological insertion for pages added from “Completed Scans.” When multiple items are picked, they are now sorted by their creation timestamp before being added to the current export session.
More…
Note: This is an RC build; feedback is welcome to improve stability and polish. - MakeACopy 2.1.0 (Release Candidate) - Rotate in cropping screen and updated workflow: you can now rotate directly on the cropping screen. - Improved OCR for mis-rotated images: the app now tries additional 90° steps (0°, 90°, 180°, 270°) and automatically selects the best result based on mean confidence (with text length as a tiebreaker). - New early-exit threshold: if the first pass already reaches a mean confidence of ≥ 85, the extra rotation attempts are skipped to deliver faster results. - Correct bounding boxes: each rotation attempt maintains an appropriate transform so word boxes are mapped accurately in the UI and export. - Stability and performance: OCR runs on a single background worker per job, with proper cancellation checks and clean Tesseract shutdown to reduce resource usage. Note: This is an RC build; feedback is welcome to improve stability and polish.
- 2.0.0 (Release) - New Scan Library: browse saved scans, basic collections, and details. - No changes to the classic export flow; existing exports continue to work as before. - Quad detection: corner angle checks now validate quads (reject overly acute/obtuse corners). - Improved candidate selection with fallback when angle constraints are violated. - OCR: selected language is now saved and restored across app launches. - Cropper: reduced interference from system edge gestures near screen borders (uses system gesture exclusion).
- 2.0.0 (Release Candidate) - New experimental Scan Library (behind a feature flag): browse saved scans, basic collections, and details. - No changes to the classic export flow; existing exports continue to work as before. - Quad detection: corner angle checks now validate quads (reject overly acute/obtuse corners). - Improved candidate selection with fallback when angle constraints are violated. - OCR: selected language is now saved and restored across app launches. - Cropper: reduced interference from system edge gestures near screen borders (uses system gesture exclusion).
- MakeACopy 1.7.6 (Release) - Fix for Xiaomi/Redmi torch darkening: On some Xiaomi/Redmi/POCO devices (e.g., Redmi Note 12 on Android 15 / HyperOS), enabling the torch made the preview/photo darker due to firmware clamping exposure. We now detect these devices and apply a positive exposure compensation automatically while the torch is on, and restore it when the torch is off or after re-binding. - Sony Xperia Android 15 camera compatibility: Enforce TextureView (COMPATIBLE) preview and conservative 4:3 preview resolution for Sony devices on Android 15+ (incl. Xperia 1 VI / 5 IV) to resolve black preview screens. - Preview watchdog timing: Increased from 1.5s to 3.0s to allow Sony lens shutters to open before re-binding the preview. - Error messages: Camera initialization/preview-failure toasts now use LENGTH_LONG to avoid truncation on some devices. - Fix OCRFragment showing outdated results on new scans by resetting OCRViewModel on new image detection. New: Added option to skip perspective correction and go straight to OCR (useful for flatbed scans or already-corrected images).
- MakeACopy 1.7.5 (Release) - Live corner preview: While scanning, a trapezoid overlay shows detected document edges in real time. - Tap-to-focus restored: Taps on the overlay now properly trigger CameraX AF/AE/AWB focus and metering. - Accessibility: Implemented performClick on the overlay so accessibility services can invoke the tap action. - Visual design improved. Smaller stability and performance tweaks while analyzing frames for the preview.
- MakeACopy 1.7.5 (Release Candidate) - Live corner preview: While scanning, a trapezoid overlay shows detected document edges in real time. - Tap-to-focus restored: Taps on the overlay now properly trigger CameraX AF/AE/AWB focus and metering. - Accessibility: Implemented performClick on the overlay so accessibility services can invoke the tap action. - Visual design improved. Smaller stability and performance tweaks while analyzing frames for the preview.
- Version 1.7.4 (Release) Fixes Black preview issue on Sony Xperia 1 VI and similar devices: Some Sony devices running Android 15 showed a black screen when opening the camera. The preview pipeline now automatically switches to TextureView (compatible mode) and uses a 16:9 preview resolution, which resolves the issue.
- Version 1.7.3 (Release) New - Added OCR languages: Portuguese (por), Dutch (nld), Polish (pol), Czech (ces), Slovak (slk), Hungarian (hun), Romanian (ron), Danish (dan), Norwegian (nor), and Swedish (swe). - Automatic locale mapping for these languages (system language → correct Tesseract code). - Updated OCR character whitelists to match each new language for better accuracy. Improvements - Added UI localization scaffolding for the new languages (strings resources), with translations where available. - Documentation and website updated to list all newly supported OCR languages.
- MakeACopy 1.7.2 (Release Candidate) - New: Low-light document detection pipeline (denoise, gamma, CLAHE, gentle sharpening). - New: Automatic ONNX output handling (supports heatmap or 8-coordinate formats via parsePrediction). - Improved: Corner accuracy with sub-pixel peak refinement and dynamic peak thresholding. - Improved: Edge detection in dark scenes by fusing Canny with adaptive low-light edges. - Improved: ONNX input preprocessing via 256×256 letterbox (no aspect distortion). - Improved: Robust point ordering around centroid and safer fallback rectangle (~10% margin). - Fixed: Resource leak in contour detection (edgesCopy) and lambda “effectively final” issue. - Fixed: Potential in-place denoising edge cases by using a temporary Mat.
More…
- Technical: Enabled opencv_photo module for FastNLMeans; replaced Mat::release method reference with a compatible loop. - Technical: Reproducible OpenCV build settings retained and compatible with changes. - MakeACopy 1.7.2 (Release Candidate) - New: Low-light document detection pipeline (denoise, gamma, CLAHE, gentle sharpening). - New: Automatic ONNX output handling (supports heatmap or 8-coordinate formats via parsePrediction). - Improved: Corner accuracy with sub-pixel peak refinement and dynamic peak thresholding. - Improved: Edge detection in dark scenes by fusing Canny with adaptive low-light edges. - Improved: ONNX input preprocessing via 256×256 letterbox (no aspect distortion). - Improved: Robust point ordering around centroid and safer fallback rectangle (~10% margin). - Fixed: Resource leak in contour detection (edgesCopy) and lambda “effectively final” issue. - Fixed: Potential in-place denoising edge cases by using a temporary Mat.
More…
- Technical: Enabled opencv_photo module for FastNLMeans; replaced Mat::release method reference with a compatible loop. - Technical: Reproducible OpenCV build settings retained and compatible with changes. - 1.7.1 (Release) - Enhance external file directory fallback for reliable storage on varied devices (SD card/internal storage). - Implement robust FileProvider URI handling with proper fallback to prevent crashes.
- 1.7.0 (Release) - Added support for larger, high-accuracy OCR models (“best” language packs). - New: Load OCR models fully offline — either via the add-on language pack or manual import of .traineddata files. - Improved OCR workflow in the app’s OCR screen to manage/download models without internet access. - Fixed: File sharing and export on SDCard-based storage (FileProvider path configuration corrected for Android 11+). Note: The add-on provides the bigger “best” models separately to keep the main app lightweight.
- MakeACopy 1.6.1 (Release Candidate) - New: Export options dialog improvements for image/PDF export. - New: Added Russian language support for OCR. - Improved: JPEG export pipeline and options for better quality and consistency. - Improved: PDF creation stability and output handling. - Improved: Black & white conversion robustness (shadow handling, adaptive thresholding). - Updated: UI texts/translations. - Technical: Minor cleanup and logging improvements. - Improved: Camera preview compatibility on Huawei/Honor devices with EMUI/Magic UI quirks.
More…
Note: This is an RC build; feedback is welcome to improve stability and polish. - 1.6.0 (Release) - Added OCR support for Chinese: Simplified (chi_sim) and Traditional (chi_tra) - Camera improvements: Auto-focus on each scan, tap-to-focus by touching the viewfinder - Enhanced image quality: Higher resolution and maximum JPEG quality for sharper scans - Accessibility: Added performClick support for better usability - Build system: Enabled parallel OpenCV Android builds for faster CI and local development - Export screen: added a small Share icon next to the Plus. It shares the last successfully exported file. - Replaced the main Share button with an Add button to capture a new page via a fresh scan (not from the registry). - Short label for the Add button (“Add”).
More…
- Trash icon is always visible; it becomes enabled only when more than one page exists. - Plus icon now jumps directly to Completed Scans (picker) without the intermediate dialog. - During saving/export, all buttons are temporarily disabled to avoid accidental actions; they are restored afterward. - Share icon is automatically enabled after a successful save/export. - Replaced the large DocAligner model with the lightweight version to reduce app size and improve performance. - ONNX build script updated: switched from CPU-only to XNNPACK and NNAPI execution providers for faster inference on Android devices. - Code cleanup and minor bug fixes - 1.5.2 (Release) - Improved reliability of OCR status updates by modernizing in‑app broadcasts (no LocalBroadcastManager; uses up‑to‑date APIs) - Stability: replaced deprecated list position APIs to ensure smoother page ordering and interactions on newer devices - Bump library versions: AGP to 8.11.2, JUnit to 1.3.0, Espresso Core to 3.7.0, Material to 1.13.0, and others. - JPEG export: fixed orientation handling so rotated scans are exported with the correct rotation (consistent with PDF export) - Preview: when "Export as JPEG" is selected, the PDF-only option "Convert B/W" is ignored; only JPEG options affect the preview
- 1.5.0 (Release) • Multi-page scanning and export (stable) - Add multiple pages from camera or "Completed scans" - Filmstrip with thumbnails; tap to preview a page - Drag-and-drop to reorder, remove single pages, or "Clear all" - Multi-page PDF with optional per-page OCR layer where available - JPEG export: single image or multi-page as ZIP (page_001.jpg, …) - TXT export combines OCR text in page order
More…
• New "Completed scans" picker - Browse previously saved scans; multi-select to add - Long-press to remove entries from the registry • Inline OCR status messages and UX polish - Improved dark mode for dialogs, better accessibility labels, and small bug fixes