Stacked 2D and Depth Sensor Integration via Molecular Bonding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging devices that combine 2D image sensors and depth sensors are either costly, bulky, or face alignment issues due to the use of different technologies and connection topologies, making it difficult to capture both 2D images and depth information effectively.
Innovation Solution
A 3D imaging device is created by superposing a 2D image sensor over a depth sensor, with the depth sensor being lit through the 2D sensor, using a silicon substrate with reduced interconnection density and molecular bonding, allowing for a single device to capture both visible light and near-infrared depth information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If separate 2D image sensor and depth sensor are used, then both 2D image and depth information can be captured, but device cost increases and device size increases
Solution Approach 1:
The patent combines a 2D image sensor and a depth sensor into a single integrated device structure. The 2D image sensor array and depth sensor array are positioned on the same substrate with overlapping fields of view, allowing both imaging functions to be captured simultaneously through a single optical path, thereby reducing device cost and size while maintaining the capability to capture both 2D images and depth information
Solution Approach 2:
The patent utilizes a three-dimensional stacked architecture where the 2D image sensor and depth sensor are positioned at different vertical levels (first level and second level) while maintaining horizontal overlap. This spatial arrangement in the vertical dimension allows both sensors to share the same optical axis and field of view, enabling integrated capture without requiring separate optical paths, thus reducing overall device complexity
2Adaptability or versatility
If separate 2D image sensor and depth sensor are used, then both 2D image and depth information can be captured, but alignment precision deteriorates due to field of view differences
Solution Approach 1:
By integrating both sensors on the same substrate with overlapping fields of view, the patent ensures that the 2D image and depth information are captured from the exact same optical perspective. This spatial merging eliminates the field of view misalignment problem that occurs with separate sensors, achieving precise automatic alignment between the two data types without requiring complex calibration procedures
3Device complexity
If 2D image pixels and depth pixels are integrated within a single array, then device compactness is improved, but manufacturing complexity increases due to different pixel technologies
Solution Approach 1:
The patent segments the sensor array into distinct regions: a first array of 2D image pixels and a second array of depth pixels, with the depth pixels positioned at a lower level. This segmentation allows each pixel type to be manufactured using its optimized technology (2D pixels for visible light imaging, depth pixels for NIR time-of-flight measurement) while maintaining overall device compactness through the stacked arrangement
Solution Approach 2:
The patent applies local quality by positioning different pixel types at different vertical levels with specialized characteristics. The 2D image pixels at the first level are optimized for visible light capture, while the depth pixels at the second level are optimized for near-infrared light detection. This local differentiation allows each region to have the specific properties needed for its function, managing manufacturing complexity through spatial separation rather than attempting to manufacture both types at the same level
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
This configuration enables a compact, cost-effective solution for capturing 2D images and depth maps, improving sensitivity to near-infrared wavelengths while reducing visible light interference, and allowing for heterogeneous technologies to be integrated within a single device.
Implementation Method 1
a depth sensor comprising a plurality of second pixels sensitive to light in the near infrared wavelength range
Implementation Method 2
a two dimensional (2D) image sensor comprising an array of first pixels
Implementation Method 3
the first and second levels are bonded together by molecular bonding
Data Source
AI summary
A three dimensional (3D) device is formed from a first level and a second level that are attached together. The first level includes a backside illuminated two dimensional (2D) image sensor including an array of first pixels sensitive to visible light. The second level includes a frontside illuminated depth sensor including an array of second pixels sensitive to near infrared light. The first and second levels are attached in a manner such that radiation, in particular the near infrared light, received at the backside of the first level passes through the first level to reach the depth sensor in the second level.