Stereo Camera Depth Calculation Using Complementary Image Enhancement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo camera modules face calibration errors and pixel mismatches due to distortion and optical aberrations, especially with short inter-lens module distances, leading to resolution deterioration and increased errors in depth calculation.
Innovation Solution
A method and device that enhance the resolution of monochromatic images using a complementary image enhancing process, involving comparison and interpolation with higher-resolution images, and applying AI-based super-resolution techniques to produce accurate depth information, utilizing monochromatic and color images captured by lens modules with different depths of field and f-numbers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the aperture is reduced to mitigate blurs, then blur is reduced, but resolution deteriorates
Solution Approach 1:
The patent changes the aperture parameter dynamically - using a large aperture for normal imaging to maintain resolution, and switching to a small aperture specifically for depth map acquisition to mitigate blur. This parameter change allows optimizing for different functions at different times
Solution Approach 2:
The patent segments the imaging function into two separate operations: one for capturing high-resolution color images with large aperture, and another for capturing depth information with small aperture. This segmentation allows each operation to be optimized independently without compromising the other
2Length of moving object
If the distance between two lens modules is reduced, then device size is reduced, but pixel mismatch increases
Solution Approach 1:
The patent introduces a third lens module as an intermediary for depth map acquisition. This intermediate lens module with small aperture provides additional depth information that helps resolve pixel mismatches between the two main lens modules, enabling accurate depth calculation even with short inter-lens distance
Solution Approach 2:
The patent replaces the mechanical solution of increasing lens module distance (to improve pixel matching) with an optical solution using aperture control and multi-lens depth mapping. By using small aperture for depth acquisition, the system achieves accurate depth measurement without needing large physical separation between lens modules
3Measurement precision
If calibration is performed for depth calculation, then depth accuracy is improved, but calibration error increases due to distortion and optical aberration
Solution Approach 1:
The patent creates a simplified optical model for depth calculation that copies only the essential geometric relationships, ignoring complex distortion and aberration effects. By using small aperture for depth acquisition, the optical path becomes simpler and更接近 to ideal pinhole camera model, reducing calibration complexity and error
Data Source
AI summary
According to an embodiment, a method for producing depth information comprises obtaining a plurality of images of an object from a plurality of lens modules, the plurality of images including at least two monochromatic images forming a monochromatic stereo image, producing two complemented monochromatic images by performing a complementary image enhancing process on at least part of the two monochromatic images, the complementary image enhancing process including comparing the plurality of images with each of the two monochromatic images and increasing a resolution of each of the two monochromatic images using an image having a higher resolution in the at least part, selecting a region of interest of the object from the two complemented monochromatic images, and calculating a depth of the region of interest by stereo-matching the two complemented monochromatic images.