1-bit Image Data Generation via Superpixel Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer programs for generating 1-bit image data from multiple-bit image data face challenges in reliable printing, as many output devices struggle to mark isolated pixels or small blocks of adjacent pixels, leading to issues like dot loss and dot gain, which result in uneven image densities and patterning.
Innovation Solution
A computer program that generates 1-bit image data by deriving 'on' and 'off' pixel values from multiple-bit image data, ensuring that each 'on' pixel value corresponds to a block of horizontally and vertically adjacent pixels, thereby improving reliability across various output devices and reducing the appearance of patterning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If isolated pixels or small blocks of adjacent pixels are used to represent dots in 1-bit image data, then the image can be printed with higher resolution and detail, but output devices cannot reliably mark these isolated pixels, causing dot loss and dot gain
Solution Approach 1:
The patent divides the image into superpixels by grouping adjacent pixels that share similar characteristics (color, intensity, spatial relationship). This segmentation transforms isolated pixels into larger, more reliable markable units while preserving image detail through the superpixel structure.
Solution Approach 2:
The patent merges adjacent pixels into superpixels by combining multiple pixel values into a single markable unit. This merging ensures that the output device marks a sufficient area to be reliably detected, eliminating dot loss and dot gain while maintaining image fidelity through the collective representation of the superpixel group.
2Productivity
If traditional AM or FM screening methods are used, then 1-bit image data can be generated, but isolated pixels or small blocks of pixels appear as patterning in the printed image
Solution Approach 1:
The patent applies local quality by making each superpixel's markable area proportional to its local image characteristics (intensity, color, importance). This allows highlights to use smaller superpixels while shadows use larger ones, eliminating uniform patterning while maintaining efficient image generation through localized adaptation.
Solution Approach 2:
The patent introduces dynamics by adaptively adjusting the superpixel size and markable area ratio based on local image content. This dynamic approach replaces static screening patterns with content-dependent superpixel structures, eliminating patterning while maintaining high productivity through automated adaptive processing.
3Reliability
If larger blocks of pixels are used to ensure reliable marking, then dot loss and dot gain are reduced, but the image resolution and detail are compromised
Solution Approach 1:
The patent segments the image into appropriately sized superpixels based on local characteristics, ensuring each superpixel is large enough for reliable marking while maintaining fine detail through dense superpixel arrangements in high-frequency areas. This segmentation achieves both reliability and resolution simultaneously.
Solution Approach 2:
The patent changes the parameter of superpixel size dynamically based on local image content, using smaller superpixels in high-detail areas and larger ones in smooth areas. This parameter adaptation maintains image resolution where needed while ensuring marking reliability where required, resolving the contradiction between resolution and reliability.
Data Source
AI summary
A computer program for, or method of, generating 1-bit image data from multiple-bit image data, by a process which comprises the steps of: receiving multiple-bit image data comprising multiple-bit pixel values; and deriving from the multiple-bit pixel values 1-bit image data comprising “on” and “off pixel values, each pixel value of the 1-bit image data corresponding to a pixel of an output medium, which pixel an output device would attempt to mark when printing the 1-bit image data if the pixel value were “on”, the 1-bit image data producing when printed an image constituted by dots, each dot corresponding to a plurality of pixel values of the 1-bit image data, which pixel values correspond to a block of M*N horizontally and/or vertically adjacent pixels of an output is medium, at least one of M and N being greater than one, wherein for at least some of the dots, where M or N is equal to one, a pixel value corresponding to a first or last pixel of a row of horizontally adjacent pixels, or to a first or last pixel of a column of vertically adjacent pixels, is “off, and where both M and N are greater than one, at least one pixel value corresponding to a pixel of a first or last row of horizontally adjacent pixels is “off and/or at least one pixel value 20 corresponding to a pixel of a first or last column of vertically adjacent pixels is “off.