Image Processing Device Edge Pixel Reference Value Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image color adjustment techniques fail to consider the color of areas other than the background, leading to issues such as light or pale colors in text and graph regions within an image during pixel value conversion.
Innovation Solution
An image processing device is employed to extract edge pixels, determine reference values, and generate correspondence data using these pixels, allowing for appropriate conversion of original pixel values into converted pixel values, thereby considering the color of areas like text and graphs distinct from the background.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If pixel values are converted based on background color from histogram, then background color adjustment is improved, but color accuracy of text and graph areas deteriorates
Solution Approach 1:
The patent segments the image into different regions by extracting edge pixels that represent text and graph areas, separating them from the background. This allows different color adjustment strategies to be applied to different regions, resolving the contradiction between background color adjustment and text/graph color accuracy.
Solution Approach 2:
The patent applies different color conversion characteristics to different regions: one characteristic for background areas and another for text/graph areas identified by edge pixels. This local differentiation ensures that background color is adjusted while preserving the color accuracy of text and graph elements.
2Illumination intensity
If correspondence data is generated using only background pixels, then background color conversion is improved, but color representation of foreground elements deteriorates
Solution Approach 1:
The patent extracts edge pixels that represent text and graph areas from the image, separating them from background pixels. This extraction ensures that the color information of foreground elements is not lost during background-oriented color conversion.
Solution Approach 2:
The patent generates different correspondence data for different regions: one set for background pixels and another set for edge pixels representing text and graphs. This ensures that color information is preserved for foreground elements while still achieving background color conversion.
3Ease of manufacture
If uniform color conversion is applied to all pixels, then processing simplicity is improved, but color accuracy of different regions deteriorates
Solution Approach 1:
The patent segments pixels into different groups (background pixels and edge pixels) and applies different conversion characteristics to each group. This maintains processing simplicity through automated classification while achieving region-specific color accuracy.
Solution Approach 2:
The patent applies different color conversion characteristics to different regions automatically based on pixel classification. This achieves region-specific color accuracy without requiring manual intervention, maintaining processing simplicity while improving color accuracy.
Data Source
AI summary
An image processing device comprises a processor and a memory. The processor executes computer-readable instructions stored in the memory. The instructions cause the image processing device to acquire target image data including each pixel value of a plurality of pixels. The instructions cause the image processing device to extract a plurality of edge pixels from the plurality of the pixels in the target image. The instructions cause the image processing device to determine a first reference value by using the plurality of the edge pixels. The instructions cause the image processing device to generate correspondence data by using the first reference value. The instructions cause the image processing device to convert, using the correspondence data, a plurality of original pixel values of the plurality of pixels into respective converted pixel values.