Digital Optical Holography for Compact 3D Surface Data Sets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for generating 3D surface data sets with depth information, such as stereoscopy and plenoptic camera systems, face limitations in resolution, computational effort, and error occurrence, especially at low light conditions, and require large device sizes or complex setups.
Innovation Solution
An optical device using digital optical holography with a dot raster and reference light to create a 3D surface data set, where relative movement between the device and object is detected to integrate successive image data sets, reducing noise and maintaining high resolution without loss of detail, enabling improved imaging quality and 'super-resolution' images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereoscopy is used to provide depth information, then depth resolution can be achieved, but a large distance between image sensors (baseline) is required and artifacts occur at edge gradients
Solution Approach 1:
The patent replaces the mechanical stereoscopy system requiring large baseline distance with a digital optical holography system that uses light interference patterns. Instead of physically separating image sensors, the system uses reference light and object light interference to encode depth information optically, eliminating the need for large mechanical baselines while achieving high depth resolution
Solution Approach 2:
The patent changes the fundamental parameter for depth measurement from spatial separation (baseline distance) to optical phase information. By measuring phase differences between reference and object light through digital optical holography, the system achieves high depth resolution without requiring large physical distances between sensing elements
2Measurement precision
If plenoptic camera systems with multi-lens optics are used, then wavefronts with different local phases can be generated, but considerable computational effort is required and errors occur at low light incidence
Solution Approach 1:
The patent extracts only the essential phase information needed for depth measurement from the full holographic data set. By using a reference light approach and focusing specifically on interference pattern analysis, the system obtains depth information with reduced computational requirements compared to full plenoptic reconstruction algorithms
Solution Approach 2:
The patent performs preliminary action by pre-establishing the reference light pattern and preparing the holographic recording geometry before actual measurement. This preliminary setup allows for more efficient real-time processing of object light interference patterns, reducing the computational burden during active measurement
3Area of moving object
If conventional photography with refractive or reflective optics is used for enlargements, then image size can be increased, but the required space and device dimensions increase
Solution Approach 1:
The patent replaces conventional refractive and reflective optical systems with digital optical holography. Instead of using physical lenses and mirrors to achieve image enlargement, the system uses optical interference and digital reconstruction algorithms, eliminating the need for large optical components and reducing overall device volume while maintaining enlargement capability
4Reliability
If multiple successive image data sets are integrated to smooth overlapping regions, then noise is reduced and signal-to-noise ratio is improved, but computational processing time increases
Solution Approach 1:
The patent performs preliminary action by pre-aligning and pre-processing successive image data sets using movement information before integration. By preparing the data sets in advance with proper registration based on detected relative movement, the system reduces the computational complexity of the integration step, achieving noise reduction with minimized processing time
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
The solution achieves high-resolution 3D surface data sets with improved signal-to-noise ratio and edge fidelity, suitable for hand-held devices, allowing for compact design and accurate depth information without significant computational overhead.
Implementation Method 1
using reference light that interferes with the raster light on the image sensor, an optical grating can be 'attached' to the object in this way
Implementation Method 2
using raster light reflected from the object and reference light internal to the casing
Implementation Method 3
object light emanating from the object is detected and a respective image data set is generated
Data Source
AI summary
The present invention relates to an optical device for examining an object (16), comprising: a housing (24); an optical unit (26), arranged in the housing (24), for incident light; a sensor unit (28) having at least one image sensor (30, 32) arranged in the housing (24); a data processing unit (34) which is coupled to the sensor unit (28) and evaluates image signals from the at least one image sensor (30, 32); an illumination unit (42), arranged at least partially on or in the housing (24), for emitting grid light toward the object (16), wherein: by means of grid light reflected by the object (16) and reference light inside the housing, a three-dimensional point grid and associated reference datasets indicative of lateral information and depth information are provided using digital optical holography; by means of the device (10; 100; 120; 140), a relative movement of the device (10; 100; 120; 140) and the objective (16) is sensed, and associated movement information in the lateral and/or depth direction is created; object light emanating from the object (16) is sensed and, in temporal succession, a respective image dataset (35) is created which is registered relative to the reference dataset in order to create a 3D surface dataset; overlapping regions of two or more successive image datasets (35) are identified on the basis of the movement information, and the overlapping regions are smoothed by integrating the associated image signals, in particular without loss of detail. The invention also relates to a method.