Capability: our OCR Omnifont Engine is indipendent from font type and size to recognize and can read virtually every kind of printed text. It reads:
Uppercase letters
Lowercase letters
Accented letters
Symbols
Punctuation
Digits
It is able to recognize characters, words, text lines, paragraphs or full pages. It's able to split and recognize glued characters and to glue and recognize broken characters, too.
Accuracy: the OCR engine accuracy is 100% when the print and image quality are good. Even if the image quality is medium, the error rate is usually very low.
Technology: our ocr recognizer is based on features extraction with multi-agent technology: for each symbol to recognize it has been predefined an agent trained to extract the features of the specific character, able to report if the input image is similar to the trained character and the similarity grade (confidence).
Speed: 200 / 600 CPS, depending on cpu speed. Using Pentium IV processors, the performance is usually over 600 CPS.
Input: Color, gray-scale o bitonal images with a 200 DPI or greater resolution can be used as input. Better performances are obtained using gray-scale image with 300 DPI resolution since the ocr engine works natively in gray-scale. To get better results, we suggest to use it with our pre-processing SDKs like deskew,
despeckle and black border removal.
Output: the output includes the ASCII or UNICODE code, the confidence level, the position and the size on the image for each recognized characters. The same informations are also available for all the recognition alternatives available for each character.
Platform: the ocr support all Windows platforms.
Packaging: the ocr engine is distributed as a singole 32 bit DLL (about a 1 MB size).
License: A SDK license is required to develop the applications. To distribute applications integrating the engine the payment of royalties can be required.
Evaluation version
Through the download section you can download an evaluation version of this product.
Plesase contact us to verify the availability of new version of SDK, in the case of problem or must particulary functionality. ADVISE: Recogniform Reader and Recogniform OCR Server are based on a different and newer OCR technology, including multiple engines voting and layout analysis features, not currently available as SDK.
Looking a solution ready to use capable of processing documents with printed or constrained characters?
Choose Recogniform Reader!
To more information about Recogniform OCR SDK and our solution for data capture and image processing, you can send us a email to informazioni@recogniform.it or fill the below form.
The library is available both in Professional and Lite version.