Stereo Camera Depth Map Accuracy via Multi-Source Segmented Lighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo cameras generate segmented depth maps due to image distortion, making it challenging to produce accurate and complete depth information.
Innovation Solution
A device and method utilizing multiple light sources with varying intensity lights to capture images, where a control unit alternates the emission of intensity lights and a processing unit generates depth information based on these images and a background image, assisting in the triangulation of accurate depth maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a stereo camera uses traditional single light source or uniform lighting to capture images for depth mapping, then the device complexity is low, but the depth map becomes segmented and distorted due to image distortion
Solution Approach 1:
The patent divides the lighting system into multiple light sources (at least three) positioned at different locations, each illuminating different regions of the imaging target. This segmentation of lighting allows capturing images from multiple lighting angles, which when processed together, produce a complete and accurate depth map without segmentation artifacts, resolving the contradiction between depth map accuracy and device complexity
Solution Approach 2:
The patent adds the spatial dimension of light source positioning by placing multiple light sources at different locations around the imaging target. This dimensional expansion enables capturing depth information from multiple perspectives simultaneously, overcoming the limitations of single-source lighting and producing undistorted, complete depth maps
2Measurement precision
If multiple light sources with different intensities are used to capture images for depth information, then the depth map accuracy improves, but the use of energy increases due to multiple light sources operating in sequence
Solution Approach 1:
The patent employs periodic action by controlling multiple light sources to emit lights with different intensities in sequence rather than simultaneously. The control unit activates each light source one after another, capturing images at different time points. This temporal sequencing reduces energy consumption compared to having all light sources operating continuously, while still gathering sufficient depth information through the series of images
Solution Approach 2:
The patent uses partial action by selecting only the necessary number of light sources (at least three) and their corresponding images to generate the depth map. Not all captured images need to be processed - only those providing essential depth information. This selective approach optimizes energy usage by avoiding redundant light source activation and image processing
Data Source
AI summary
A device for generating depth information includes a plurality of light sources, an image capture unit, a control unit, and a processing unit. Each light source of the plurality of light sources emits a corresponding intensity light. The control unit controls the each light source of the plurality of light sources to emit the corresponding intensity light in turn, and controls the image capture unit to capture a plurality of images corresponding to the plurality of light sources. The processing unit generates a plurality of depth information corresponding to the plurality of light sources according to the plurality of images corresponding to the plurality of light sources and a background image after the image capture unit captures the plurality of images corresponding to the plurality of light sources and the background image.