Adaptive autocolor detection based on media color

Abstract

A method for detecting background color of a scanned document uses dynamic weights to adjust components of the L*a*b* color space. In accordance with one aspect of the disclosure, the L*a*b* values belonging to pixels within a block are applied to a series of thresholds. In response to the L*a*b* values meeting the thresholds, a given color pixel within the block is reclassified as being monochromatic. The weight values are dynamically updated based on at least one of the L*a*b* values of the given pixel. The thresholding is repeated for remaining pixels in the document. Using the classified pixels, an output document is generated where the background pixels are monochromatic. The output document is transmitted to a display device in communication with the server computer.

Description

BACKGROUND[0001]The present disclosure relates to digital printing. It finds particular application in conjunction with detecting color objects in a scanned image and will be described with particular reference thereto. It will be appreciated, however, that the invention is also amenable to other like applications.[0002]At times, it is desirable to differentiate neutral monochromatic (e.g., gray) pixels from color pixels in an image. For example, electronic documents on color media or color background may want to be stored or printed on monochrome background, without effecting the color content, such as images, graphics and text. Auto color detection is used to classify document type as gray or color. The existing approach for differentiating pixels is referred to as auto color detection. Auto color detection is a block-based approach that operates on a reduced L*a*b* image.[0003]In other words, the conventional approach for detecting the background color of an original scanned imag...

AI Technical Summary

Benefits of technology

This patent is about a method for detecting the background color of a scanned document by using a computer system. The system applies a series of thresholds to a block of pixels around a given pixel, and if the pixels meet the thresholds, they are classified as being monochromatic. The system dynamically updates predetermined weight values based on the colors of the pixels. This allows the system to generate an output document with a monochrome background. The technical effect of this patent is an improved method for extracting background color information from scanned documents, which can be useful in various applications such as document management and information retrieval.

Problems solved by technology

This classification may result in detection failures, particularly when documents are printed on recycled media.

Industries such as financial institutions suffer from this problem because they routinely handle forms on recycled media.

As a result, the colored content (e.g., text, print) can become more difficult to distinguish in a scanned image that is detected in color.

The text is obscured in some places or difficult to read.

Images