Radio Tomographic Image Formation via Provincial Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radio tomographic imaging systems face significant computational complexity as the number of sensors increases, leading to prohibitive computational time and challenges in real-time monitoring.
Innovation Solution
A new algorithm that partitions the imaging area into provinces, allowing each province to be solved independently, reducing the computational burden by inverting matrices fewer times and using a provincial least squares method to generate radio tomographic images efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of sensors in the network is increased to improve image quality and coverage, then measurement precision and area of stationary object are improved, but computational complexity increases leading to prohibitive computational time
Solution Approach 1:
The patent divides the imaging area into multiple provinces, allowing the computational problem to be segmented into smaller sub-problems. Each province can be processed independently or in parallel, reducing the overall computational complexity while maintaining image quality. This segmentation approach enables the system to handle larger sensor networks without proportionally increasing computational time.
2Area of stationary object
If the number of sensors in the network is increased to improve image quality and coverage, then area of stationary object is improved, but device complexity increases
Solution Approach 1:
By dividing the large imaging area into smaller provinces, the system can manage complex sensor networks more effectively. Each province represents a manageable computational unit that can be processed independently, reducing the overall system complexity while maintaining comprehensive area coverage.
3Measurement precision
If contemporary RTI methods are used to produce better imagery, then measurement precision is improved, but computational complexity increases making real-time monitoring difficult
Solution Approach 1:
The provincial decomposition method enables parallel processing of different regions, significantly improving processing throughput and enabling real-time monitoring capabilities while maintaining good image quality.
Solution Approach 2:
The patent applies regularization techniques that focus computational effort on the most relevant aspects of the problem, achieving good enough solutions more quickly than exhaustive methods would allow.
Data Source
AI summary
A method for generating radio tomographic images is provided. A plurality of transceivers positioned around a region to be imaged is divided into a plurality of pixels. A control apparatus is configured to cause each of the plurality of transceivers in turn to send a signal to each of the other transceivers. The control apparatus is further configured to determine an attenuation in the received signals, generate weighing, derivative, and attenuation matrices from the signals, group the pixels into a plurality of provinces, select each province in turn and solve for a change in attenuation in each of the pixels while setting the pixels in other provinces to zero, aggregate solutions from each of the provinces into a rough estimate, re-solve each province using the aggregated rough estimate, aggregate the re-solved solutions from each province into a refined estimate, and generate an image from the refined estimate.