Depth Map Refinement for iToF Phase Unwrapping Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for obtaining depth maps using indirect Time of Flight (iToF) sensors face issues with inconsistent distance calculations due to errors in phase unwrapping or weak light signal strength, leading to inaccuracies in the depth map.
Innovation Solution
A method involving a window-based classification and scanning process to identify and correct or discard image pixels with inconsistent distances, using confidence factors and distance thresholds to refine the depth map.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If indirect time of flight measurement is used to obtain depth map, then depth information can be acquired, but inconsistent distance calculations occur due to phase unwrapping errors or weak light signals
Solution Approach 1:
The patent applies local quality by analyzing each pixel's neighborhood window independently to detect and correct local inconsistencies. Each pixel is evaluated based on its specific local context (neighboring pixels), allowing localized correction of phase unwrapping errors or weak signal issues without affecting the entire depth map. This localized approach improves reliability while maintaining measurement precision.
Solution Approach 2:
The patent implements feedback by using the depth values of neighboring pixels to verify and potentially correct the distance calculation of each pixel. The consistency check mechanism provides feedback on whether a pixel's measured distance is reliable, and if not, it uses information from surrounding pixels to correct the inconsistency, thereby improving overall measurement reliability.
2Measurement precision
If phase unwrapping is performed to calculate distance, then depth information is obtained, but errors in phase unwrapping lead to inconsistent distance calculations
Solution Approach 1:
The patent applies preliminary anti-action by performing a consistency check on the phase unwrapping results before finalizing the depth map. By evaluating whether the calculated distance is consistent with neighboring pixels and meeting confidence criteria, the system preemptively identifies and corrects potential phase unwrapping errors before they propagate through the entire depth map, thereby maintaining both precision and reliability.
3Measurement precision
If light reflected signal strength is too weak compared to ambient light, then depth measurement becomes unreliable, but increasing signal strength may not be feasible
Solution Approach 1:
The patent uses feedback by calculating a confidence factor for each pixel based on the strength of the reflected light signal relative to ambient light. When the signal is too weak, the confidence factor indicates low reliability, triggering a consistency check that uses neighboring pixel information to correct or discard the unreliable measurement, thereby maintaining overall measurement reliability without requiring increased signal strength.
Solution Approach 2:
The patent introduces an intermediary mechanism (the consistency check based on neighboring pixels) that mediates between weak signal measurements and reliable depth map output. When a pixel's signal is too weak to be reliable, the intermediary process uses information from surrounding pixels to infer the correct depth value, effectively bridging the gap between weak signals and reliable measurements.
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
Improves the accuracy of depth maps by identifying and correcting or removing inconsistent distance measurements, resulting in a more reliable and consistent depth map output.
Implementation Method 1
each imager pixel of a pixel array of an indirect time of flight sensor acquire samples during a phase of acquiring the scene to be imaged during which the scene is irradiated with a frequency-modulated light signal. For each imager pixel, a phase shift between the signal irradiating the scene and the signal reflected from the scene and received by the pixel is determined from the samples acquired by the pixel, and a distance is determined from this phase shift.
Data Source
AI summary
A method comprises, for each pixel of a depth map, acquiring samples and calculating a distance, and defining a window of N*N pixels with this image pixel at the center, N being an odd integer equal to or greater than 3, and, for each window, classifying the pixels of the window into groups, based on a threshold and distances calculated, calculating for each group, a number of pixels in the group, a confidence factor of the group equal to a sum of the confidence factors of the pixels each determined from the samples acquired for the pixel, and scanning the window from the central pixel by comparing, for each pixel, the confidence factor of the group of the pixel with a threshold and the number of pixels of the group of the pixel with another threshold, and determining whether the central pixel is retained, replaced or discarded.