Wide Dynamic Range Depth Imaging via Time-Multiplexed Structured Light
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Structured light imaging devices face challenges in accurately estimating the three-dimensional depth of scenes due to variations in object reflectance, leading to under-exposure of darker objects and over-exposure of brighter objects, which results in undetectable projected patterns and poor quality depth maps.
Innovation Solution
A method and apparatus for wide dynamic range depth imaging using a structured light imaging device that projects a structured light pattern at varying projection times and captures images at corresponding exposure times, generating multiple depth images which are then combined to create a composite depth image, thereby improving depth map quality across a wide range of albedo values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed camera exposure time and projector projection time are used, then the device complexity is reduced, but darker objects become under-exposed and brighter objects become over-exposed, making the projected pattern undetectable in these regions
Solution Approach 1:
The patent applies dynamics by making the exposure time and projection time adjustable rather than fixed. The system dynamically adapts the exposure time based on the reflectance properties of different objects in the scene, allowing optimal capture of both dark and bright regions. This is achieved through multiple depth images captured at different exposure times, which are then combined to produce a high-quality depth map across the entire dynamic range.
Solution Approach 2:
The patent changes the exposure time parameter to resolve the contradiction. By capturing multiple depth images at different exposure times and combining them, the system achieves accurate depth estimation for both under-exposed dark objects and over-exposed bright objects, effectively expanding the measurable dynamic range without increasing device complexity.
2Measurement precision
If multiple depth images are captured at different projection times and exposure times, then the depth map quality across wide albedo range is improved, but the loss of time increases
Solution Approach 1:
The patent uses partial action by capturing multiple depth images at different exposure times only for the regions that require it. The system identifies under-exposed and over-exposed regions and selectively processes them with appropriate exposure times, rather than uniformly processing the entire scene. This reduces the overall time penalty compared to capturing multiple images of the entire scene at all exposure levels.
Solution Approach 2:
The patent creates multiple copies of the depth image at different exposure times and combines them. Instead of capturing one perfect image, the system captures several partial copies (depth images at different exposures) and synthesizes a complete high-quality depth map by merging the valid regions from each copy, efficiently achieving comprehensive coverage.
3Illumination intensity
If the camera exposure time is increased to capture darker regions, then darker objects are better exposed, but brighter objects become over-exposed and the pattern becomes undetectable
Solution Approach 1:
The patent segments the scene into different reflectance regions (dark, medium, bright) and applies different exposure times to capture each segment optimally. By dividing the scene based on albedo properties and processing each segment with appropriate exposure parameters, the system ensures that dark regions are properly exposed without over-exposing bright regions, maintaining pattern detectability across all segments.
Solution Approach 2:
The patent changes the exposure time parameter based on the local reflectance characteristics of different regions. For dark regions, a longer exposure time is applied to capture sufficient light, while for bright regions, a shorter exposure time is used to prevent over-exposure. This localized parameter adjustment maintains reliable pattern detection across the full dynamic range of the scene.
Data Source
AI summary
Wide dynamic range depth imaging in a structured light device is provided that improves depth maps for scenes with a wide range of albedo values under varying light conditions. A structured light pattern, e.g., a time-multiplexed structured light pattern, is projected into a scene at various projection times and a camera captures images of the scene for at least the same exposure times as the projection times. A depth image is computed for each of the projection/exposure times and the resulting depth images are combined to generate a composite depth image.