Ground-Penetrating Radar Image Coregistration for Multi-Pass Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial GPR systems face challenges in accurately coregistering images from multiple passes due to imprecise navigation sensors and signal interruptions, making it difficult to combine images effectively and enhance weak features by reducing background noise.
Innovation Solution
A system and technique for coregistering images by preprocessing, feature extraction, and matching common features between image blocks, followed by transforming them to a common coordinate system and combining them to form a larger image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If navigation sensors are used to coregister images from multiple passes, then image combination is enabled, but imprecision and calibration errors degrade alignment accuracy
Solution Approach 1:
The patent introduces an intermediary feature-matching process that mediates between the GPR images and navigation sensors. By extracting and matching features (such as reflections from subsurface objects) directly within the image data, the system creates a reference framework that compensates for navigation sensor imprecision and calibration errors, thereby improving alignment accuracy while maintaining multi-pass image combination capability
Solution Approach 2:
The patent implements a feedback mechanism where the matched features from multiple passes provide corrective information back to the coregistration process. The feature matching results are used to refine and adjust the alignment between images, compensating for errors in navigation sensor data and progressively improving alignment accuracy across multiple passes
2Loss of information
If images from multiple passes are combined, then weak features can be enhanced and noise reduced, but coregistration difficulty increases due to sensor imprecision
Solution Approach 1:
The patent divides the coregistration problem into manageable segments by processing images in pairs or small groups from multiple passes. By segmenting the overall coregistration task into smaller, more manageable alignment operations between individual passes, the system reduces the complexity of coordinating multiple imperfect navigation datasets while still achieving the cumulative benefit of enhanced weak features and noise reduction through combination
3Measurement precision
If feature extraction and matching is performed to improve alignment, then coregistration precision increases, but processing complexity and time increase
Solution Approach 1:
The patent applies partial action by performing feature extraction and matching on only the most salient or distinctive features within the GPR images, rather than processing all possible features. This selective approach extracts sufficient information to achieve accurate coregistration while significantly reducing the computational time and processing resources required compared to exhaustive feature analysis
Data Source
AI summary
A method of coregistering images is disclosed. Two dimensional images are first preprocessed, including clipping values in each image to a specified standard value range, and then applying a local equalization technique to the clipped images. Feature extraction is then performed to identify one or more features in the images. One or more of the extracted features are then matched between images to determine a degree of overlap between the images and a relationship between image coordinates of the images. The images are then transformed to a common coordinate system based on the degree of overlap and the relationship between their image coordinates. The images are then combined to form a single image.