SAR Image Analysis Using Weighted Parameter Likelihoods
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image analysis methods using synthetic aperture radar (SAR) technology struggle with noise variability in electromagnetic wave reflections, leading to unreliable elevation and displacement analysis results due to multiple peaks in evaluation values, making it difficult to determine the true optimal values.
Innovation Solution
An image analysis device and method that utilize complex matrix calculation, parameter candidate selection, candidate evaluation using a weight matrix, and statistics calculation to assess the reliability of analysis results by reducing noise effects and evaluating parameter likelihoods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing image analysis methods using SAR technology are used, then elevation and displacement analysis can be performed, but noise variability in electromagnetic wave reflections causes multiple peaks in evaluation values, reducing reliability of results
Solution Approach 1:
The patent segments the evaluation process by dividing parameter candidates into multiple groups and performing separate evaluations for each group. This segmentation allows the system to identify the true optimal value by comparing results across different segments, thereby resolving the issue of multiple peaks caused by noise and improving both reliability and measurement precision.
Solution Approach 2:
The patent introduces an intermediary evaluation process that assesses the reliability of parameter candidates before final selection. By adding this intermediary step that calculates reliability based on evaluation values and compares them across candidate groups, the system can filter out false peaks caused by noise and identify the true optimal parameter value, thus improving reliability without sacrificing precision.
2Reliability
If multiple peaks appear in evaluation values due to noise, then it becomes difficult to determine the true optimal parameter values, but increasing noise filtering may reduce the accuracy of parameter estimation
Solution Approach 1:
The patent implements a feedback mechanism where the reliability of parameter candidates is calculated based on their evaluation values, and this reliability information is fed back into the selection process. The system uses this feedback to iteratively refine parameter selection, comparing candidates across different groups and adjusting selections based on reliability metrics, thereby determining true optimal values while maintaining estimation accuracy.
Solution Approach 2:
The patent changes the approach from direct parameter selection to a multi-stage process where parameters are evaluated, ranked by reliability, and selected based on comparative analysis across candidate groups. By transforming the selection criterion from raw evaluation values to reliability-based ranking, the system can distinguish true optimal values from noise-induced peaks while preserving measurement precision.
3Reliability
If complex matrix calculation and candidate evaluation methods are used to reduce noise effects, then computational complexity increases, but this is necessary to improve the reliability of analysis results
Solution Approach 1:
The patent segments the complex computational process into distinct modules: complex matrix calculation, candidate evaluation, reliability calculation, and parameter selection. By organizing the computation into segmented stages with clear interfaces, the system manages computational complexity while maintaining the necessary calculations for reliable results. Each segment performs a specific function, making the overall complex process more manageable and efficient.
Solution Approach 2:
The patent performs partial evaluations by dividing parameter candidates into multiple groups and evaluating each group separately rather than evaluating all candidates simultaneously. This partial action approach reduces the computational burden of the complex matrix calculations while still achieving reliable results through comparative analysis across the partial evaluations, thus balancing reliability with manageable computational complexity.
Data Source
AI summary
The image analysis device 100 includes a complex matrix calculation unit 11 which calculates a complex matrix that reflects a phase difference in all pairs of images in multiple images in which a same region is recorded, a parameter candidate selection unit 12 which selects multiple candidates of parameter which explains a phase shift, a candidate evaluation unit 13 which evaluates likelihood of the multiple candidates using the complex matrix and a predetermined weight matrix, and a statistics calculation unit 14 which weights the candidates of parameter by the likelihood and calculates statistics of the candidates of parameter.