Dual Readout Sensor for Fourier Transform Profilometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fringe-projection profilometry techniques, such as Fourier transform profilometry, face challenges in achieving accurate three-dimensional reconstruction due to non-uniform intensity and reflectance distributions, particularly in low-cost fringe projectors and human subjects with varying reflectance, leading to inaccuracies in depth measurement.
Innovation Solution
A system that uses a dual readout sensor to capture reflections of both frequency and zero-frequency signal patterns, or phase-shifted frequency patterns, and performs DC subtraction or cancellation to generate a wrapped phase map, which is then unwrapped for three-dimensional reconstruction, addressing the non-uniformity issues and improving accuracy and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Fourier transform profilometry is used for three-dimensional reconstruction, then depth measurement capability is provided, but measurement precision deteriorates due to non-uniform intensity and reflectance distributions
Solution Approach 1:
The patent extracts and removes the DC component (zero-frequency signal) from the captured fringe pattern using a dual readout sensor. By separating and eliminating the DC component that causes non-uniform intensity artifacts, the remaining AC component contains only the pure fringe information needed for accurate phase calculation and three-dimensional reconstruction
Solution Approach 2:
The patent introduces a DC readout channel as an intermediary element between the fringe projector and the phase calculation process. This intermediary channel specifically captures the DC component, allowing it to be subtracted from the main fringe pattern before Fourier transform analysis, thereby eliminating its harmful effects on measurement precision
2Measurement precision
If conventional single readout sensor is used, then system complexity is kept low, but measurement precision deteriorates due to inability to handle non-uniform reflectance
Solution Approach 1:
The patent segments the sensor readout function into two independent channels: an AC readout channel for capturing fringe patterns and a DC readout channel for capturing zero-frequency signals. This segmentation allows each channel to be optimized for its specific function and enables precise removal of DC artifacts without requiring complex post-processing algorithms
Solution Approach 2:
The patent implements a dual readout sensor that performs multiple functions simultaneously: it captures both the AC fringe component and the DC component in a single integrated device. This multi-functional sensor eliminates the need for separate sensors or complex mechanical modulation systems, achieving high measurement precision while maintaining relatively simple system architecture
3Measurement precision
If multiple frames are captured for DC subtraction, then measurement precision improves, but productivity deteriorates due to increased processing time
Solution Approach 1:
The patent performs DC component capture and subtraction as a preliminary action during the image capture phase itself, using the DC readout channel to record the zero-frequency signal that is then immediately subtracted from the AC fringe pattern. This preliminary removal of DC artifacts eliminates the need for iterative processing or multiple frame captures, achieving both high precision and fast reconstruction speed
Data Source
AI summary
A system for three-dimensional object sensing using fringe-projection profilometry with Fourier transform analysis is described. A fringe-projection profilometry (FPP) projector simultaneously transmits a frequency signal pattern (i.e., modulated or AC signal) and a zero-frequency signal pattern (i.e., unmodulated or DC signal) onto an object's surface. Alternatively, the projector projects two frequency signal patterns phase-shifted by 180 degrees. A dual readout sensor captures reflections of both signals from the object's surface as adjacent frames, and the DC signal is extracted by subtraction of addition to obtain an enhanced signal. The enhanced signal is used to generate a wrapped phase map through Fourier transform profilometry (FTP). The resulting wrapped phase map is unwrapped, and three-dimensional reconstruction of the object's surface is generated by converting phase from the unwrapped phase map to three-dimensional coordinates.