Microscope Distance Determination Using Machine Learning Overview
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining the distance of a sample reference plane in microscopy are inefficient, requiring multiple images or additional cameras, and are not suitable for varying sample carrier heights, leading to potential collisions and focusing issues.
Innovation Solution
A method using a single overview image captured by a microscope camera, processed by an evaluation unit with a trained machine learning model, to detect characteristics of the sample carrier and determine the distance from a reference plane, eliminating the need for multiple images or additional cameras.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple overview images are taken from different angles using triangulation, then distance determination accuracy is improved, but device complexity increases due to requiring two overview cameras or defined relative movement
Solution Approach 1:
The patent extracts the distance determination function from complex multi-camera triangulation systems and implements it through a single camera combined with machine learning analysis of overview images. The evaluation unit processes single images to detect sample carrier characteristics and determine distance, eliminating the need for multiple cameras or complex mechanical movement systems.
Solution Approach 2:
The patent replaces the mechanical/optical triangulation system with an information processing system based on machine learning. Instead of using multiple cameras and geometric triangulation, a single camera captures overview images that are analyzed by an evaluation unit using trained machine learning models to directly determine distance and sample carrier characteristics.
2Adaptability or versatility
If sample carriers of different heights are used, then versatility is improved, but reliability deteriorates due to potential collisions and focusing issues
Solution Approach 1:
The patent performs preliminary distance determination and sample carrier characterization using machine learning analysis of overview images before the actual microscopy measurement begins. The evaluation unit detects characteristics such as sample carrier type, height, and position in advance, allowing the system to pre-adjust parameters and avoid collisions during subsequent operations.
Solution Approach 2:
The patent implements a feedback mechanism where the evaluation unit continuously analyzes overview images to determine sample carrier characteristics and distance, then uses this information to adjust system parameters. This closed-loop approach ensures reliable operation with varying sample carrier heights by constantly monitoring and adapting to the actual sample carrier configuration.
Data Source
AI summary
A method for determining a distance of a sample reference plane of a sample carrier from a reference plane of a microscope, the microscope including a sample stage for the sample carrier and a camera, comprises the following steps: taking an overview image of the sample carrier by means of the camera; evaluating the overview image and thus detecting at least one characteristic of the sample carrier; ascertaining contextual data of the characteristic from a data set; and determining the distance of the sample reference plane from the reference plane based on the characteristic and the contextual data of the sample carrier. A microscope configured to determine the distance of the sample reference plane of the sample carrier from the reference plane is also described.