Non-uniform laminar pattern`s e.g. book, two-dimensional image equalizing method, involves assigning two-dimensional image point in scanned image to each determined three-dimensional surface point of pattern

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention is a method for digitally 'flattening' or equalizing images of non-uniform, patterned surfaces (such as pages in a book or textiles) that might be curved or distorted. It works by scanning the surface with a moving light stripe and capturing image sequences, then reconstructing the exact three-dimensional shape. Each spot (point) on the 3D surface is matched to a corresponding point in a 2D scanned image, which is then used to generate a corrected, undistorted image of the original pattern.

Use CasesContent extracted from patent full text and abstract with AI.

• Digitally reproducing pages from rare or delicate books without physical flattening.
• Scanning and archiving curved or folded artworks, textiles, or documents.
• Quality control in manufacturing for inspecting patterns on curved surfaces.
• Restoring or digitizing historical documents and artifacts with complex surface geometry.
• Automated digitization in libraries, museums, and archives of materials that cannot be pressed flat.

BenefitsContent extracted from patent full text and abstract with AI.

• Accurately removes distortions from scanned images of uneven or non-flat surfaces.
• Preserves delicate or valuable items by eliminating the need for physical flattening during scanning.
• Enables high-fidelity digital archiving and reproduction of complex three-dimensional patterned surfaces.
• Improves the quality and legibility of digitized documents and artifacts.
• Can be integrated with automated scanning systems for efficient and large-scale digitization projects.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The method involves producing a scanned image during a scanning process of a scanning device e.g. line camera (1). A light stripe is moved over a pattern (4), and image sequences of the pattern are recorded by a web camera (3). A three-dimensional surface point of the pattern is determined from the image sequences by temporal and spatial evaluation of the moved light stripe. A two-dimensional image point in the scanned image is exactly assigned to each determined three-dimensional surface point of the pattern, and an equalized original two-dimensional copy of the pattern is computed.