Underwater Robot Vision Using Color Shift for 3D Point Cloud
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot vision systems for underwater environments rely on active sensors like lasers or structured light, which require emission and are costly, whereas a passive method using digital imaging to estimate relative distances based on color shifts in captured images is advantageous for generating a 3D point cloud without emitting light.
Innovation Solution
A robot vision apparatus comprising digital cameras, a camera filter, and a processor that captures unfiltered and filtered images, determines pixel matches, and calculates image distances using the Beer-Lambert law to generate a 3D point cloud by comparing RGB values, thus estimating attenuation and relative distances without requiring knowledge of initial light intensity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If active sensors like lasers or structured light are used for underwater robot vision, then measurement precision and reliability are improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces active mechanical/optical sensing systems (lasers, structured light projectors) with a passive digital imaging system that uses standard cameras and computational algorithms. The mechanical complexity of laser ranging systems is substituted with software-based image processing that analyzes color attenuation in natural light images to infer distance information.
Solution Approach 2:
The system uses the ambient light already present in the underwater environment rather than requiring additional active illumination. The natural light serving the primary function of illuminating the scene also provides the color attenuation information needed for distance estimation, eliminating the need for separate ranging hardware.
2Illumination intensity
If active light emission is used for underwater imaging, then image quality and visibility are improved, but energy consumption and system complexity increase
Solution Approach 1:
The system utilizes ambient light that is already present in the underwater environment for both illumination and distance measurement. No additional energy is expended to generate illumination, as the natural light serving the primary function of illuminating the scene also provides the color attenuation information needed for distance estimation.
3Measurement precision
If multiple cameras and filters are used to capture unfiltered and filtered images, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The imaging system is segmented into distinct functional components: an unfiltered camera for capturing overall scene information and a filtered camera for capturing wavelength-specific attenuation information. This segmentation allows each component to be optimized for its specific function while keeping individual component complexity low.
Solution Approach 2:
The colored filter acts as an intermediary that selectively transmits certain wavelengths of light while blocking others. This intermediary component enables the system to extract distance information through color attenuation without requiring complex multi-spectral imaging systems, as the filter converts wavelength information into visible color differences that can be processed by standard cameras.
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
Enables the generation of a 3D point cloud representing the surrounding environment by passively sensing relative distances through color shifts in underwater environments, enhancing navigation capabilities without the need for active light emission, thus reducing costs and complexity.
Implementation Method 1
A filtered image is captured using a camera filter which tends to pass certain wavelength bandwidths while mitigating the passage of other bandwidths
Implementation Method 2
determines an image distance for the match pair using the unfiltered and filtered first color space coordinates and an absorption coefficient value
Data Source
AI summary
A robot vision system for generating a 3D point cloud of a surrounding environment through comparison of unfiltered and filtered images of the surrounding environment. A filtered image is captured using a camera filter which tends to pass certain wavelength bandwidths while mitigating the passage of other bandwidths. A processor receives the unfiltered and filtered images, pixel matches the unfiltered and filtered images, and determines an image distance for each pixel based on comparing the color coordinates determined for that pixel in the unfiltered and filtered image. The image distances determined provides a relative distance from the digital camera to an object or object portion captured by each pixel, and the relative magnitude of all image distances determined for all pixels in the unfiltered and filtered images allows generation of a 3D point cloud representing the object captured in the unfiltered and filtered images.