We currently have Kimberly Stoutjesdijk working with us as an intern on evaluation of an open-source OCR engine, and have an internal working group consisting of colleagues from the Research, Collections, Digitisation, Information policy and DBNL departments tasked with researching the quality and possible correction of our digital collections.
We hope to further develop Ochre for post-correction on microfilm scans and plan to use the ground-truth developed in this project for more evaluation projects. The dataset of ground-truth files from our digitised newspaper collection is available for research purposes.
Please contact dataservices kb. If you are interested in an internship using our digital collection, ground-truth datasets or OCR quality, please send us a message to discuss possible topics and options. NL Search. Development of GT Once we established that we could in fact use the access images for our research instead of the hard to access master images we selected pages from the collection. Type what you see including capitals and printing mistakes if there are any.
Any layout information, such as bullet points or indentations, or style, such as bold or italics, can be ignored. Only type logos and other textual lines in more visual parts of the newspapers such as advertisements if they are clearly legible.
Initials are part of the first line. Ignore spaces in words that are there for layout reasons. Post-correction with machine learning: Ochre We worked together with research software engineer dr. The word error rate is calculated as follows, where the order of the words is not taken into account for the order independent option: Fig 1.
Description of calculation of word error rate by Carrasco, Text Digitisation Results Since we subdivided the selection into time periods and OCR software the results can be displayed for various subsets. Lessons learned One of the major lessons we learned during this project is that ground-truth production is always more complicated and takes more time than what you plan for.
Lotte Wilms. Digital Scholarship Advisor. Some files, printed nearly years ago such as Latvian newspapers and magazines utilizing Gothic font needed to converted to electronic files.
Responsible for identifying large documents, OCR will be an important part of digital library projects. OCR technology is an important part of any document management system, in which OCR is mainly used to recognize characters in an image to reduce manual entry time.
The below problems often occur during recognition. About Contact Privacy Policy. You may have crash issues on latest operating systems. If your file is reversed, you can rotate to have a more accurate OCR. Also, it supports spell check, you can replace those suspected errors with words from the dictionary. Yet, it may take you some time to manually adjust the errors. Though it is designed to convert files to editable Word, the formatting cannot keep in the Word file.
In addition, the last update for this program was released in , there may be some technical errors on different, especially latest Windows. Anyway, it offers high accuracy and deserves a try. In addition, it offers a set of image and document tools, to view and manage the file easily, which gives a better user experience. Does this program look familiar to you? I have been hesitate if I should put this SuperGeek program on this list, anyway, it is an option to do OCR for free and offline, for users who have problems accessing to Free OCR to Word, this program may help.
OneNote from Microsoft is a note-taking program to create and organize notes across different platforms, including macOS, Windows, Android and iOS, also it offers online free portal to manage your notes anytime anywhere.
0コメント