Sonar Image Stabilization via Model Distortion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sonar imaging technologies face challenges in visualizing and accurately imaging sparse data from surfaces immersed in fluids, particularly due to low resolution and artifacts such as multipath reflections, which distort the image and make it difficult to fit models to sonar data effectively.
Innovation Solution
The method involves using sonar pulses to record reflected signals, calculating data to produce a sonar image, and then adjusting either the sonar or model data to account for distortions, allowing for more accurate fitting of a model to sonar data by inflating or deflating the model image to match the sonar image's resolution and distortions, thereby stabilizing the image from ping to ping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sonar imaging is performed with standard resolution, then the imaging process is simple and fast, but the image quality is poor with low resolution and artifacts
Solution Approach 1:
The patent applies preliminary action by pre-distorting the model data to match expected sonar artifacts before fitting. The model is pre-inflated or pre-warped based on anticipated distortion patterns, so that when the distorted sonar image is received, the pre-adjusted model aligns more accurately with the actual sonar data, improving resolution without requiring complex real-time correction algorithms
Solution Approach 2:
The patent changes parameters by adjusting the resolution and distortion characteristics of the model data to match the sonar imaging parameters. The model's geometric parameters are modified (inflated/deflated) to compensate for known sonar artifacts, allowing the fitting process to achieve higher effective resolution while maintaining computational simplicity
2Measurement precision
If model fitting is performed on distorted sonar images, then the fitting accuracy improves, but the fitting stability from ping to ping deteriorates due to image distortion variations
Solution Approach 1:
The patent implements feedback by using the fitted model from one ping to inform and stabilize the fitting process for subsequent pings. The successfully fitted model parameters serve as a reference frame that constrains and guides the fitting of distorted images from successive pings, preventing drift and maintaining temporal stability while still accommodating distortion variations
Solution Approach 2:
The patent applies preliminary action by establishing a reference model fit from initial pings before attempting to fit distorted images from subsequent pings. This preliminary fitted model serves as a stable reference framework that helps anchor the fitting process, reducing sensitivity to distortion variations and improving temporal consistency across multiple pings
3Measurement precision
If the model data is inflated to compensate for sonar artifacts, then the alignment with sonar image improves, but the model accuracy relative to the actual object decreases
Solution Approach 1:
The patent applies segmentation by dividing the model into different regions or components that can be selectively adjusted. Different parts of the model are inflated or deflated by different amounts based on their specific distortion characteristics, allowing the model to align with sonar artifacts in some regions while maintaining geometric accuracy in other regions where distortion is minimal
Solution Approach 2:
The patent implements local quality by applying non-uniform inflation or deflation to different regions of the model based on local distortion patterns. The model's geometric properties are modified locally rather than globally, so that alignment with sonar artifacts is achieved in distorted regions while preserving accuracy in regions where the sonar imaging is less affected by artifacts
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables the production of more accurate and stable 3D sonar images, allowing for precise tracking and placement of objects by compensating for sonar imaging artifacts, improving the alignment of model and sonar data, and reducing visual jittering effects.
Implementation Method 1
directing a sonar pulse at the object and recording sonar signals reflected from the object
Implementation Method 2
directing a sonar pulse at the object and recording sonar signals reflected from the object with a sonar imaging array
Data Source
AI summary
A method of tracking a known object is presented, wherein a sonar image of an object which is distorted by an artifact associated with sonar imaging is compared with an image generated from a model of the object, and at least one of the two images is modified to reduce differences between them.