Image Processing Device Shadow Restoration via Resolution Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies fail to accurately restore shadows or shielding objects in images due to differences in camera positions and angles, leading to unnatural results when trying to match and paste parts from secondary images with lower resolution.
Innovation Solution
An image processing device and method that acquires both primary and secondary images, identifies edit target areas, adjusts the secondary image's resolution to match the primary image, and edits the primary image based on similar areas to improve restoration accuracy, including the use of surrounding similar areas and estimation units for enhanced detail and color fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the second image is photographed from a position farther than the first position or at a different angle, then the shadowed part or shielded part can be seen in the second image, but the photographed objects appear differently in size and shape between the first and second images
Solution Approach 1:
The patent applies dynamics by making the image processing adaptive to different camera positions and angles. The system dynamically adjusts the photographed part area in the second image through deformation processing to match the shape and size of the edit target area in the first image, rather than using a fixed processing approach. This allows the restoration to account for perspective changes while maintaining accuracy.
Solution Approach 2:
The patent changes parameters by adjusting the resolution and deformation of the photographed part area from the second image to match the edit target area in the first image. Specifically, the system decreases the resolution of the photographed part area and applies deformation processing to transform its shape and size parameters, enabling accurate restoration despite differences in camera position and angle.
2Loss of information
If the second image is used to restore the shadowed part or shielded part, then information about the hidden area can be obtained, but the resolution decreases due to editing operations
Solution Approach 1:
The patent applies local quality by performing resolution adjustment and deformation processing only on the photographed part area that needs to be restored, rather than processing the entire second image. This localized approach minimizes quality loss while obtaining the necessary information about the shadowed or shielded part. The system identifies the specific photographed part area and processes only that region to match the edit target area.
3Ease of operation
If simple pasting of the photographed part area from the second image is applied, then the restoration process is simple, but the restoration accuracy is insufficient due to resolution decrease and unnatural appearance
Solution Approach 1:
The patent applies preliminary action by performing resolution decrease and deformation processing on the photographed part area before pasting it onto the first image. This preprocessing ensures that the photographed part area is properly prepared and matched to the edit target area in terms of resolution, shape, and size, thereby improving restoration accuracy while maintaining operational simplicity through automated processing.
4Shape
If the photographed objects are enlarged or deformed to match between first and second images, then the images can be aligned, but the resolution decreases due to editing operations
Solution Approach 1:
The patent applies segmentation by dividing the second image into a photographed part area and an unphotographed part area, and further dividing the first image into an edit target area and a non-edit target area. This segmentation allows the system to process only the relevant photographed part area through resolution decrease and deformation, rather than processing the entire image, thereby minimizing overall resolution loss while achieving proper alignment.
Data Source
AI summary
An image processing device is configured to: acquire a first image photographed from a first position; acquire a second image photographed from a second position; identify an edit target area where a first shadow or a first shielding object is shown in the first image; decrease a resolution of the second image so as to match with a photographed object shown in the first image, and identifying a photographed part area where a second shadowed part or a second shielded part is shown in a low-resolution second image; identify a similar area in at least one of the first image and the second image, the similar area being similar to the photographed part area; and edit the edit target area in the first image, based on the similar area.