Released: Jan 15, 2021

V21.1 中的更新

特性

• Added the ability to save recognition result in plain text, Microsoft Word, Office Open XML or ODF Text Document format.

修補程式

• Fixed bug with length of result lines.

Released: Nov 29, 2020

V20.11 中的更新

特性

• Added the ability to recognize a batch of images in a zip archive or in a folder. Pass archive name or folder name as a parameter and get an array of results from each page.

Released: Oct 22, 2020

V20.10 中的更新

特性

• Added the option to set a custom angle for skew correction.
• Extended API: Added the ability to choose language options.

Released: Sep 25, 2020

V20.9 中的更新

特性

• Added new JSON recognition result format.
• Added new API method:
  • Send URI-link for recognition.
  • Send recognition options as an object.
  • Get recognition result as an object with different properties.

Released: Aug 31, 2020

V20.8 中的更新

特性

• Added a new Machine Learning (ML) model with improved symbol recognition.

Released: Jul 31, 2020

V20.7 中的更新

特性

• Added the ability to remove vertical and horizontal lines from tables.
• Added image recognition using GPU and CUDA technology (available in a separate assembly Aspose.OCR.Gpu).

Released: Dec 19, 2017

V17.11 中的更新

特性

• New Binary Classification provides better OCR quality.
• New Neural Network Player (NNNPlayer) provides better OCR performance.

Released: Jul 27, 2017

V17.6 中的更新

特性

• Added hOCR support for the creation of searchable PDFs.
• Improved logic to help set the space between words.
• Added Clusterizator to reduce the number of calls to OCR and use average images for OCR.
• Outlines and clustering splitters to separate touching letters.

Released: Apr 5, 2017

V17.03 中的更新

特性

• The metered licensing mechanism has been improved. The API will switch to unpaid mode if it fails while uploading consumption data. The Custom HostnameVerifier has been removed.
• Processing of deskewing skewed images has been improved.

Released: Mar 2, 2017

V17.02 中的更新

特性

• Improved gaussian filter mechanism.
• Improved upper and lower case character recognition mechanism.

修補程式

• Incorrect line splitter behaviour processing lines with 'i' letter.    
• Exception within sort rectangles during layout detection.