Temporal Compressive Sensing for TEM Frame Rate
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current implementations of coded-aperture video compressive sensing in transmission electron microscopes (TEMs) face substantial practical limitations due to the difficulties in producing and maintaining the coded aperture, which results in invasive and costly system modifications, and may not achieve the desired frame rate increase due to physical constraints like contamination and limited resolution.
Innovation Solution
The proposed method employs temporal compressive sensing by capturing multiple full-resolution images per data acquisition period, each representing a distinct linear combination of time slices, using a high-speed switching system to direct different linear combinations of radiation patterns to various regions of a sensor array, allowing for reconstruction of time slice datasets with a time resolution exceeding the data acquisition period.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If coded-aperture video compressive sensing is implemented in TEMs, then frame rate increase is achieved, but system complexity and cost increase due to invasive modifications and maintenance difficulties
Solution Approach 1:
The invention extracts the coded aperture functionality from the physical optical path and relocates it to the temporal domain through software-based compressive sensing algorithms. This eliminates the need for physical coded aperture masks and their associated mechanical switching systems, thereby reducing system complexity while maintaining frame rate enhancement capabilities
Solution Approach 2:
The invention replaces the mechanical coded aperture system with a computational approach. Instead of physically modulating light through masks and shutters, the system uses temporal sampling and mathematical reconstruction algorithms to achieve compressive sensing, eliminating mechanical complexity and maintenance requirements
2Loss of information
If coded aperture is used, then compressive sensing is achieved, but resolution and signal quality deteriorate due to contamination and physical constraints
Solution Approach 1:
The invention creates multiple temporal copies of the signal through rapid sequential sampling at different time points. By capturing multiple frames and combining them through compressive sensing reconstruction, the system achieves high signal quality and resolution without the degradation associated with physical coded apertures
Solution Approach 2:
The system employs periodic temporal sampling with carefully designed sampling intervals and patterns. This periodic action in the temporal domain enables compressive sensing reconstruction while maintaining signal quality and resolution, avoiding the contamination and degradation issues of physical masks
3Loss of time
If multiple images are captured per data acquisition period, then time resolution improves, but data processing complexity increases
Solution Approach 1:
The system performs preliminary temporal sampling and organizing of multiple frames before reconstruction. By pre-processing the temporal data and structuring it according to the compressive sensing sampling pattern, the computational complexity of the reconstruction phase is reduced while maintaining high time resolution
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 significant improvements in time resolution and data acquisition rates, overcoming the limitations of spatially-encoded methods by directly capturing multiple images per block of time, reducing computational complexity, and minimizing blurring artifacts, while maintaining high signal utilization efficiency.
Implementation Method 1
capturing sensor array data for one or more data acquisition periods, wherein within each of the one or more data acquisition periods, one or more measurement datasets corresponding to distinct linear combinations of patterns of the radiation transmitted, reflected, elastically scattered, or inelastically scattered by the sample or scene are captured
Data Source
AI summary
Methods and systems for temporal compressive sensing are disclosed, where within each of one or more sensor array data acquisition periods, one or more sensor array measurement datasets comprising distinct linear combinations of time slice data are acquired, and where mathematical reconstruction allows for calculation of accurate representations of the individual time slice datasets.