Histogram-Based Gamma Curve Generation for Image Density Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing apparatuses face challenges in accurately judging the type of an image and performing proper density adjustments, often resulting in improper adjustments due to erroneous judgments, especially when dealing with images having dark backgrounds or varying gradation widths.
Innovation Solution
An image processing apparatus that calculates histograms, detects highlight and shadow parts, and produces a γ curve based on both, allowing for accurate density adjustments tailored to the image by considering both highlight and shadow parts for improved judgment and adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a γ curve is selected based on only the highlight part or shadow part of the histogram, then the density adjustment can be performed automatically, but the image type judgment becomes erroneous leading to improper density adjustment
Solution Approach 1:
The histogram is segmented into multiple regions: highlight part (high luminance), shadow part (low luminance), and half-tone part (intermediate luminance). Each region is analyzed separately to extract characteristic values, which are then combined to produce a comprehensive γ curve that accurately represents the entire image distribution, resolving the contradiction between automation and judgment accuracy.
Solution Approach 2:
Different γ values are determined for different regions of the histogram based on their local characteristics. The highlight part uses one set of characteristic values, the shadow part uses another, and the half-tone part uses a third. This local differentiation allows each region to contribute its specific quality to the overall γ curve, improving both automation reliability and judgment accuracy.
2Device complexity
If the density adjustment focuses only on the distribution of the histogram, then the processing is simplified, but the judgment is incorrect for images with varying gradation widths
Solution Approach 1:
The histogram distribution is segmented into distinct regions (highlight, shadow, half-tone) with different luminance characteristics. By analyzing each region separately and combining their characteristic values, the system achieves precise density adjustment for images with varying gradation widths without excessive processing complexity.
Solution Approach 2:
The system changes the parameter approach from using a single distribution metric to using multiple region-specific characteristic values (foot value, peak value, start point for each region). This parameter expansion enables precise density adjustment while maintaining reasonable processing complexity through automated extraction and combination of these parameters.
Data Source
AI summary
The present invention relates to a technique implementing a density adjustment fitted to an image, concretely, an image processing apparatus for (1) judging the kind of a read image with a high accuracy, and (2) producing an arbitrary γ curve. The apparatus includes: a histogram calculation section for calculating a histogram of an image; a highlight part detection section for detecting a highlight part from the histogram; a shadow part detection section for detecting a shadow part from the histogram; a γ curve production section for producing a γ curve by using γ values corresponding to each of the highlight part and the shadow part, both obtained on detection results of the highlight part detection section and the shadow part detection section, respectively; and a density adjustment section for performing a density adjustment of an image on the γ curve produced by the γ curve production section.