Depth Map Quality Scoring via Flatness and Jaggedness Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Depth maps generated from images often contain inaccuracies due to factors like lighting, obstructions, subject movement, and device calibration issues, leading to flat regions, jagged edges, and patchy areas, which affect their quality and reliability in applications such as facial recognition.
Innovation Solution
A method to analyze depth maps for flatness, jaggedness, and patchiness by calculating specific metrics, normalizing them, and determining a quality score, which can trigger calibration processes or quality labeling to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth maps are generated from images using algorithms or sensors, then depth information can be obtained for various applications, but inaccuracies such as flat regions, jagged edges, and patchy areas appear in the depth map
Solution Approach 1:
The patent implements a feedback mechanism by calculating quality metrics (flatness, jaggedness, patchiness) from the depth map and using these metrics to determine whether recalibration is needed. The system continuously monitors depth map quality and feeds this information back to trigger recalibration when quality degrades, thereby resolving the contradiction between measurement precision and reliability.
Solution Approach 2:
The patent performs preliminary recalibration actions based on predicted quality degradation. By calculating quality metrics in advance and comparing them against thresholds, the system triggers recalibration before severe inaccuracies affect application performance, maintaining both measurement precision and reliability.
2Reliability
If recalibration is performed to improve depth map quality, then future depth maps can be more accurate, but the calibration process consumes time and resources
Solution Approach 1:
The patent applies partial recalibration by performing quality assessment on specific regions of the depth map (flat regions, edges, patchy areas) rather than requiring full recalibration. The system calculates quality metrics selectively and triggers recalibration only when necessary and only for affected regions, reducing calibration time while maintaining depth map quality.
Solution Approach 2:
The patent changes the parameter being measured from raw depth values to quality metrics (flatness, jaggedness, patchiness). By transforming the depth map into quality metric maps, the system can efficiently assess whether recalibration is needed without processing the entire depth map at full resolution, reducing calibration time while maintaining reliability.
3Measurement precision
If quality metrics are calculated for all regions of the depth map, then comprehensive quality assessment is achieved, but computational complexity increases
Solution Approach 1:
The patent segments the depth map into three distinct quality metric maps: flatness metric map, jaggedness metric map, and patchiness metric map. Each metric focuses on specific types of inaccuracies, allowing the system to assess quality comprehensively while keeping individual calculations computationally efficient. This segmentation enables targeted quality assessment without requiring complex full-map analysis.
Data Source
AI summary
Methods are provided to determine a quality score for depth map. The quality score is calculated from metrics that detect artifacts or other inaccuracies in the depth map such as flat patches, artifactual edges, and patchy regions. A flatness metric detects regions of neighboring pixels that have substantially the same depth value. A jaggedness metric detects hard edges or other discontinuities. A patchiness metric detects regions that are wholly enclosed by an edge and that have sub-threshold areas. The individual metrics are normalized and combined to determine an overall quality score for the depth map. The quality score can then be compared to one or more thresholds to determine a quality label for the depth map. Such a quality label can then be used to unlock a device, to invalidate an unlock attempt, to recalibrate a depth sensor, or to perform some other operations.