The nature of the Arabic language is imposing difficulties; revealed to variance of shapes, lack of standards of fonts, unlimited scope of irregularities that stood against the existence of such a product. Thus, it has been one of the century challenges - in the field of image processing - to develop an Arabic OCR product that is: reliable, practical, flexible, user friendly and with high accuracy. With ICRA Gold, ASSET has proved to be one of the market leaders in this field.

ICRA GOLD introduces an Arabic OMNI OCR engine that recognizes any text without the need for teaching the system any font, or taking any samples for shapes. The OMNI engine adopts new technologies depending upon contextual information besides the geometrical nature of the font. Having an OMNI OCR engine - for recognizing Arabic text - did not deprive the user from the trainability features. Changing the fonts through various training tools is made to enhance the recognition accuracy.

ICRA GOLD provides:

• Language building capabilities which make ICRA capable of supporting multiple languages, not just Arabic
• Recognition of Arabic documents using a built-in OMNI base engine
• Font-learning capabilities that can work as a process in parallel with the recognition operation
• A 32-bit application that utilizes multi-threading processing capabilities, hence, users will gain the edge of multiprocessing and recognition for multiple images
  
  

Some of ICRA Gold Features:

• Integrated scanning capabilities
• Powerful image enhancement functions including Deskew, Despekle, Invert, Rotate, Binary Filters, Erosion , Dilation
• Save Image with different commonly known formats
• Support for diactrics
• Cutting/Segmentation options leading to higher recognition quality
• Filtering Graphics
• Preserving horizontal and vertical indentation of recognized text
• Recognizing multiple images at the same time
• Defining categories of documents (Computer output, typewrites, magazines, newspaper...etc.) that will eventually enhance the recognition accuracy