Automatic Background Removal Using Dynamic LUTs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing background suppression techniques in image processing struggle with inconsistent results due to variations in devices and media, leading to artifacts like 'punch-through' in halftone and highlight regions, especially with poor contrast between foreground and background.
Innovation Solution
An image processing system that converts input RGB image data to L*a*b* data, processes the background by determining a white point, mapping entries to one-dimensional look-up tables, and performing trilinear interpolation based on the mapped data, while also conducting post-process weighting and background adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If threshold-based segmentation algorithms are used to suppress background, then background uniformity is improved, but abrupt switching artifacts (punch-through) appear in halftone and highlight regions
Solution Approach 1:
The patent applies parameter changes by using multiple thresholds (first threshold and second threshold) instead of a single threshold value. The first threshold is used for initial background identification, and the second threshold is applied to prevent over-correction in halftone and highlight regions. This multi-threshold approach modifies the parameter (threshold value) to balance background uniformity with artifact prevention.
2Manufacturing precision
If background suppression is applied to improve contrast between foreground and background, then background uniformity is improved, but foreground regions may be incorrectly modified
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different image regions through multiple thresholds. The first threshold handles general background areas, while the second threshold specifically protects halftone and highlight regions. This localized approach ensures that background suppression is applied uniformly where appropriate while preserving foreground integrity in sensitive regions.
3Manufacturing precision
If complex background suppression algorithms are used to handle poor contrast images, then background removal accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies segmentation by dividing the background suppression process into distinct stages with different thresholds. The first threshold handles the primary background suppression task, while the second threshold addresses specific challenging regions (halftone and highlight areas). This segmented approach breaks down the complex problem into manageable steps, improving accuracy without requiring excessively complex algorithms that would increase processing time.
Data Source
AI summary
A computationally efficient way is provided to enable background adjustment based on real time, automatic background detection. In one form, the real time detected paper white is used to dynamically adjust 1D LUTs that map the input pixel values before accessing a 3D LUT that stores the weights between the original input pixel value and white (e.g., 255, 128, 128 in L*a*b*). The more complex calculation of generating coefficients in the 3D LUT is made independent of the paper white point. In one form, the white point is calculated based on the histogram of the image collected from the lead edge or the full page.