Image Alignment Apparatus Using Dual Conversion Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image alignment technologies face challenges in accurately aligning images from different sensors, particularly in scenarios where feature point detection is difficult, leading to issues with real-time alignment and image fusion in monitoring and medical imaging applications.
Innovation Solution
An image alignment apparatus and method that estimates a first conversion function based on feature point information and a second conversion function based on motion information, allowing for selection of the appropriate conversion function for real-time alignment, using feature point and motion vector detection and similarity determination units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature point detection is used for image alignment, then alignment accuracy is improved, but reliability deteriorates when feature points are difficult to detect
Solution Approach 1:
The system changes the parameter basis for alignment from feature point information to motion information. When feature points are difficult to detect, the system switches to using motion vectors extracted from temporal changes in image data, thereby maintaining alignment reliability under varying detection conditions while preserving accuracy through appropriate parameter selection
Solution Approach 2:
The system dynamically selects between two different conversion functions based on detection conditions. The conversion function selection unit determines whether to use a feature point-based conversion function or a motion-based conversion function according to the ease of feature point detection, making the alignment system adaptive to different scenarios and improving overall reliability
2Adaptability or versatility
If multiple conversion functions are estimated and selected, then adaptability is improved, but device complexity increases
Solution Approach 1:
The system employs dynamic selection between two conversion functions based on real-time assessment of feature point detectability. The conversion function selection unit evaluates detection conditions and switches between feature point-based and motion-based conversion functions, providing adaptability to different scenarios without requiring a completely complex system redesign
Solution Approach 2:
The conversion function selection unit acts as an intermediary that manages the complexity of having multiple conversion functions. It evaluates detection conditions and automatically selects the appropriate conversion function, thereby providing adaptability while keeping the system architecture manageable through a centralized selection mechanism
Data Source
AI summary
Provided are an image alignment apparatus and an image alignment method using the same. The image alignment apparatus includes a first conversion function estimation unit for estimating a first conversion function based on feature point information that is extracted from a first image, which is captured by using a first image sensor, and a second image, which is captured by using a second image; and a second conversion function estimation unit for estimating a second conversion function based on motion information that is extracted from the first image and the second image.