Image Processing with OpenCV and OCR

Task 1: Image Extraction

• From Manuscript source: doc/docx/ppt/pptx/xlx/xlsx

• Identify type (Raster, Vector) and Colorspace (Grayscale/Color) of the images

• Standalone images: Raster (jpg, png, gif etc.), Vector(eps, ai, svg etc.)

Task 2: Machine Learning Based Classification algorithms for OCR (standard text font )

• Produce mark-up (text enclosed by rectangles) for verification by user

• Text Enhancement

• Identify level of blur

• Identify font size

• Perform presets (Gaussian blurring/un-sharp mask/threshold/level)

Task 3: Image Sizing & Colorspace Mapping

• Use a set of rules on the input image resolution and identified text to interpolate and produce high resolution images

• Output the images in RGB/Grayscale/CMYK based on rules provided

NOTE: Image Processing Libraries like OpenCV &/or other OCR libraries can be used to accomplish these tasks

Detailed requirements will be shared with selected vendors.

Image Processing Machine Learning (ML) Оптичне розпізнавання символів OpenCV

ID Проекту: #16101397

Про проект