Image Correction for Moving Object Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques face challenges in generating high-quality images by correctly detecting and removing moving objects, particularly due to positional misalignments and blurs caused by factors like camera shake and moving trees, and struggle to accurately determine whether areas with crowds or constantly moving objects are background or foreground, leading to ghosting issues.
Innovation Solution
An image processing device and method that sets a correction target area in an image based on multiple input images, using one of the images as a correction image to correct the target area, and determines the reliability of the correction area to decide on necessary corrections, employing a reliability determining unit and appearance changing area detection to handle moving objects and blurs effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a background image is generated by calculating a weighted average of multiple input images, then moving objects can be removed from the image, but small positional misalignments and blurs occur due to hand-shaking and moving objects like tree leaves
Solution Approach 1:
The patent divides the image into multiple regions and processes each region separately. By segmenting the image and identifying foreground objects in each segment, the system can apply different processing strategies to different areas, thereby reducing the impact of misalignment and blur on the overall image quality while maintaining effective moving object removal.
Solution Approach 2:
The patent applies local quality by treating different regions of the image differently based on their characteristics. Foreground regions with moving objects are processed with higher priority and different parameters compared to background regions. This allows the system to maintain high precision in critical areas while tolerating minor misalignments in less critical areas.
2Reliability
If areas with crowds or constantly moving objects are used as background, then the background can be determined, but such areas cannot be properly corrected and ghosts appear in the image
Solution Approach 1:
The patent employs feedback mechanisms to continuously evaluate and refine the identification of foreground and background regions. By analyzing multiple input images and comparing regional characteristics across frames, the system can detect when an area should be classified as foreground rather than background, thereby preventing ghost artifacts while maintaining accurate background determination.
Solution Approach 2:
The patent applies dynamics by making the foreground/background classification adaptive and time-varying. Instead of static classification, the system dynamically adjusts region labels based on observed motion patterns across multiple frames. This allows constantly moving objects to be correctly identified as foreground and excluded from background processing, preventing ghosting while maintaining reliable background determination.
Data Source
AI summary
The present disclosure is to generate a high-quality image by correcting a predetermined correction target image based on a plurality of input images. In an image processing device 3, an image correcting section 160 detects a user tap gesture on a touch panel 250. When the position of the tap gesture is within foreground candidate areas detected by a foreground candidate area detecting section 140, the image correcting section 160 corrects a base image set by a base image setting section 120 in the areas corresponding to the foreground candidate areas.