Pixelnetica Document Imaging SDK provides developers with an intelligent, highly efficient toolkit, which offers an easy way to add image processing features that are optimized for document photos that are taken by a mobile device or document camera to their applications.
Designed to ensure the smooth operation of a paperless workflow by pre-processing paper document images, SDK makes them easier to handle by text recognition (or optical character recognition - OCR) programs, enhancing the visual quality and legibility of documents.
Short video demonstrates how Document Imaging SDK works in real application - SharpScan.
«Smart Crop» feature worked out and Black & White profile applied.
Want to try?Request Information and Free Trial
Document Imaging SDK main Features and Benefits
- Optimized for Speed and Quality for each supported platform, especially crucial for mobile.
- All operations run in Full Automatic mode - without the necessity of user interference.
- Precise document area boundaries detection.
- “Smart Crop” - offers the possibility to estimate the quality of the document area boundary detection and make a decision about cropping – crop image automatically or send for user approval and correction.
- “Smart Camera” estimates the document scan quality and provides user with live feedback during document shooting to to reduce common user scanning errors.
- 2D geometric and 3D perspective distortion (e.g. trapezoid) correction.
- Document image align and rotation.
- Camera matrix digital noise reduction.
- Brightness and contrast equalization including strong shadows mitigation.
- Document background cleaning.
- Adaptive image binarization optimizes for future text recognition (OCR).
- Content oriented adaptive Color processing profile automatically generates perfect documents.
- Special presets for producing Grayscale and Color documents optimized for legibility.
Improve the OCR Quality of mobile Document Images
Paper documents digitized by traditional scanners are usually well lit, almost perfectly oriented, and benefits from scanners of high speed and duplexing support.
Document images taken by mobile cameras usually suffer from the human factor (shooting errors, shaking hands, bad lighting, shadows, glare etc.) so specific image processing has to be applied to get the best achievable OCR results:
- 2D & 3D geometric distortions should be removed;
- Images should be much more rotated (comparable to those from a scanner);
- Digital noise produced by the camera matrix should be properly detected and eliminated;
- Shadows, glare, and document backgrounds need to be removed;
- Documents should be carefully binarized (converted to Black & white) to save the maximum details for OCR.
The absence of such specific image preprocessing in turn leads to a huge error level during text recognition, and frequently making it impossible.
Pixelnetica Document Imaging SDK provides all the necessary preprocessing features to achieve the best possible OCR results.
Difference from traditional imaging SDK’s
Most well-known imaging SDK aimed for ordinary photo manipulations like: image format processing for reading and writing, simple image rotation, brightness and contrast changing, general purpose binarization, etc.
However, this is not enough for paper document photos, especially if they are meant to be OCRed (processed with optical character recognition system to convert image to text).
Simply taking a picture of a document with a camera (especially a mobile device camera) does not result in document quality images. Images display geometric distortion resulting from the angle of the camera in relation to the document and often have unwanted background shadows.
Pixelnetica Document Imaging SDK was created to overcome these inherent shortcomings and to produce high-quality documents just as if they were scanned with a conventional scanner.
How it works in real application
Most of Document Imaging SDK’s features have been integrated into Pixelnetica SharpScan – mobile document scanner that allows users to scan documents and photos on the go.
Ready to get started?
Please get in touch to request free trial SDK, additional information or price quotation.
Trial SDK is similar to commercial except of results watermarking and includes documentation as well as sample application source code.