3D Particle Localization via Point Spread Function Dictionary
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current optical imaging systems face limitations in resolving the number and location of particles, especially when they are densely packed, as they are constrained by the diffraction limit, requiring multiple frames and compromising temporal resolution for super-resolution imaging.
Innovation Solution
The method employs a point spread function dictionary to estimate coefficients using algorithms like Matching Pursuit and Convex Optimization, allowing for the determination of particle number and location, even in densely packed scenes, by decomposing images into coefficients corresponding to particle positions and omitting noise, enabling super-resolution imaging with fewer frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional optical imaging systems are used to image densely packed particles, then the system is constrained by the diffraction limit, but multiple frames are required to achieve super-resolution, compromising temporal resolution
Solution Approach 1:
The patent pre-computes a dictionary of point spread function responses for all possible particle positions before imaging. This preliminary preparation enables rapid coefficient estimation during actual imaging without requiring multiple frames, achieving super-resolution in a single frame while maintaining high temporal resolution
Solution Approach 2:
The patent replaces the conventional mechanical/optical approach of capturing multiple frames with a computational approach using dictionary-based coefficient estimation. This substitution allows super-resolution to be achieved through mathematical decomposition rather than temporal averaging, preserving temporal resolution
2Measurement precision
If multiple frames are collected to achieve super-resolution imaging, then measurement precision improves, but data collection time and computational effort increase
Solution Approach 1:
The patent extracts only the essential information (point spread function coefficients) directly from the captured image through dictionary matching, rather than collecting multiple frames and processing them. This extraction approach achieves super-resolution from a single frame, significantly improving data collection speed while maintaining precision
Solution Approach 2:
The patent changes the imaging approach from temporal averaging (multiple frames) to spatial decomposition (dictionary coefficient estimation). By transforming the problem into estimating coefficients for pre-computed point spread function responses, it achieves super-resolution with single-frame imaging, improving productivity
3Measurement precision
If conventional imaging methods are used for dense particle scenes, then the diffraction limit constrains resolution, but increasing frames improves measurement precision
Solution Approach 1:
The patent introduces a pre-computed dictionary of point spread function responses as an intermediary between the captured image and the particle localization result. This dictionary serves as a reference library that enables direct coefficient estimation from a single frame, eliminating the need for multiple frames and reducing system complexity
Solution Approach 2:
The patent performs preliminary computation of the point spread function dictionary before actual imaging. This pre-computation creates a ready-to-use reference that simplifies the imaging process, allowing particle number and location to be determined directly from single-frame images without requiring multiple frames
Data Source
AI summary
Systems, methods, and computer program products are disclosed to localize and/or image a dense array of particles. In some embodiments, a plurality of particles may be imaged using an imaging device. A plurality of point spread function dictionary coefficients of the image may be estimated using a point spread function dictionary; where the point spread function dictionary can include a plurality of spread function responses corresponding to different particle positions. From the point spread function dictionary coefficients the number of particles in the image can be determined. Moreover location of each particle in the image can be determined from the point spread function dictionary coefficients.