Composite Image Metric for Star Pattern Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pattern recognition technologies face challenges such as ambiguity, false identification, and increased complexity due to measurement errors and field-of-view limitations in applications like fingerprint, iris, and star pattern recognition, leading to misidentification and increased computational requirements.
Innovation Solution
The method involves computing a composite image metric from multiple digitized images using metrics like separation distance, area, or central angles, and comparing it to a reference metric to improve pattern recognition accuracy, reducing measurement errors and increasing the effective field of view by combining data from multiple images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data from multiple star images is combined and averaged to reduce measurement error radius, then star position measurement precision is improved, but device complexity and computational requirements increase
Solution Approach 1:
The patent segments the star recognition process into multiple independent image captures taken at different orientations. Each image is processed separately to extract star positions and compute metrics, then the results are combined. This segmentation allows measurement error reduction through averaging while managing computational complexity by processing images in discrete, manageable units rather than as a single complex operation.
Solution Approach 2:
The patent performs preliminary actions by capturing multiple star images at different spacecraft orientations before the final pattern recognition decision. These pre-captured images with their associated metrics are stored and then combined through averaging to reduce measurement errors. This preliminary data collection and processing reduces the computational burden during the critical recognition phase while improving measurement precision.
2Reliability
If the effective field of view is increased to include more stars for unambiguous pattern recognition, then recognition reliability is improved, but device complexity increases due to polynomial or factorial increase in memory or computation usage
Solution Approach 1:
The patent adds a temporal dimension to the star recognition process by capturing images at multiple different spacecraft orientations rather than relying on a single wide-field image. This dimensional change allows the system to achieve unambiguous pattern recognition through combining metrics from multiple viewpoints, effectively increasing the usable star field without requiring a single large, computationally intensive field-of-view capture.
Solution Approach 2:
The patent merges data from multiple star images taken at different orientations by combining their respective metrics through averaging. This merging process creates a composite view that includes more stars and provides unambiguous pattern recognition while managing computational complexity by using efficient combination algorithms rather than processing a single large dataset.
3Reliability
If measurement error radius is reduced through reduced star position measurement errors, then ambiguity and false identification are reduced, but measurement precision requirements become more stringent
Solution Approach 1:
The patent implements feedback by comparing star position measurements from multiple images and using the consistency of these measurements across different orientations to reduce the effective measurement error radius. The repeated measurements provide feedback that allows the system to identify and eliminate outliers, thereby reducing ambiguity and false identification while achieving the required precision through statistical validation rather than demanding ultra-precise single measurements.
Data Source
AI summary
In an exemplary embodiment, a pattern is recognized from digitized images. A first image metric is computed from a first digitized image and a second image metric is computed from a second digitized image. A composite image metric is computed as a function of the first image metric and the second image metric, and a pattern is identified by comparing the composite image metric against a reference image metric. The function may be a simple average or a weighted average. The image metric may include a separation distance between features, or a measured area of a feature, or a central angle between two arcs joining a feature to two other features, or an area of a polygon whose vertices are defined by features, or a second moment of a polygon whose vertices are defined by features. The images may include without limitation images of friction ridges, irises, or stars.