Image Processing Apparatus Edge Strength Background Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques for document data, such as photo books, fail to optimize background image selection for character objects, leading to reduced visibility due to edge strength differences between the background and character objects, making it difficult to distinguish them effectively.
Innovation Solution
An image processing method that identifies the edge strength of background regions in candidate images and selects images with smaller edge strength values to ensure optimal visibility of superimposed character objects, using edge detection filters like Sobel filters and calculating edge strength evaluation values to determine suitable background images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional image selection methods are used that only consider image quality evaluation criteria, then the image selection process is simple, but the visibility of superimposed character objects is reduced due to edge strength differences
Solution Approach 1:
The patent performs preliminary edge strength evaluation on candidate background images before final selection. By calculating edge strength values in advance for each candidate image and comparing them against threshold values, the system prepares evaluation data that guides the subsequent image selection process, ensuring character visibility is maintained from the outset.
Solution Approach 2:
The patent introduces edge strength as a new evaluation parameter alongside traditional image quality metrics. By changing the selection criteria to include edge strength comparison against threshold values, the system transforms the image selection process from considering only image quality to considering both image quality and character visibility, thereby resolving the contradiction.
2Manufacturing precision
If images with high edge strength are selected as background images, then the image quality may be high, but the distinguishability between character objects and background is reduced
Solution Approach 1:
The patent applies local quality evaluation by specifically assessing the edge strength in the background region where character objects will be superimposed. Instead of evaluating the entire image uniformly, the system focuses on the local background area's edge characteristics, allowing high image quality overall while ensuring the specific region where characters appear has suitable edge properties for distinguishability.
Solution Approach 2:
The patent changes the evaluation parameter from general image quality alone to a combination of image quality and local edge strength in the background region. By introducing edge strength as a comparative parameter against threshold values, the system enables selection of images that maintain both high quality and adequate character distinguishability.
3Reliability
If edge strength evaluation is added to the image selection process, then character visibility is improved, but the processing complexity increases
Solution Approach 1:
The patent performs preliminary edge strength evaluation on candidate images to establish baseline metrics before final selection. By calculating edge strength values in advance and comparing them against pre-determined threshold values, the system prepares evaluation data that streamlines the subsequent selection process, maintaining efficiency while improving character visibility.
Solution Approach 2:
The patent replaces subjective visual assessment of character visibility with automated edge strength calculation and threshold comparison. This substitution of mechanical/computational analysis for human judgment improves both the objectivity of evaluation and the efficiency of the selection process, as automated calculations can be performed rapidly without manual review.
Data Source
AI summary
An image processing method to edit document data including an image on which a character object is superimposed, includes: obtaining a candidate image group that includes one or more candidate images each being a candidates of an image to be arranged in the document data; identifying, for each candidate image, as a background region a region that is a background of the superimposed character object in a case where the candidate image is arranged in the document data; detecting an edge of the candidate image in the background region and identifying a value related to an edge strength based on the detected edge; and selecting as a selected image a candidate image with the value related to the edge strength of the background region that is smaller than a predetermined threshold.