Light-Field-Binocular Depth Calculation via Confidence Map Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing depth detection methods, such as light field cameras and binocular cameras, face challenges in obtaining accurate depth information across a full range, with light field cameras struggling at far distances due to small baselines and stereo cameras facing limitations at close distances due to sparse angle resolution.
Innovation Solution
A light-field-binocular system is constructed using a monocular camera and a light field camera, calibrating images to generate input images for algorithms, obtaining disparity maps based on binocular and light field information, combining confidence maps to optimize and convert into a final depth map, allowing for accurate depth estimation across various distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a light field camera is used for depth detection, then angle resolution is improved, but baseline is reduced leading to poor far-distance depth measurement
Solution Approach 1:
The patent combines a light field camera and a binocular camera into a hybrid system that leverages the strengths of both technologies. The light field camera provides high angle resolution for near-distance objects, while the binocular camera provides adequate baseline for far-distance depth measurement. The system fuses depth maps from both cameras to achieve comprehensive depth coverage across all distances.
2Length of moving object
If a stereo camera with large baseline is used for depth detection, then far-distance depth measurement is improved, but angle resolution is reduced leading to poor close-distance depth measurement
Solution Approach 1:
The patent merges the capabilities of binocular and light field cameras to resolve this contradiction. The binocular camera's large baseline is utilized for far-distance depth measurement, while the light field camera's high angle resolution compensates for the limited close-distance performance. The fused depth map integrates results from both systems to provide accurate depth information across the full distance range.
3Device complexity
If only binocular camera is used, then system complexity is reduced, but depth measurement accuracy across full depth range is insufficient
Solution Approach 1:
The patent segments the depth measurement task into two parts: near-distance measurement handled by the light field camera and far-distance measurement handled by the binocular camera. This segmentation allows each camera to operate in its optimal performance range, improving overall depth measurement accuracy while maintaining reasonable system complexity through modular architecture.
Data Source
AI summary
A depth information calculation method and device based on a light-field-binocular system. The method includes obtaining a far-distance disparity map based on binocular information of calibrated input images, setting respective first confidences pixels in the disparity map, and obtaining a first target confidence; detecting the first confidence of a pixel being smaller than a preset value and responsively determining a new disparity value based on light field information of the input images, determining an update depth value based on the new disparity value, and obtaining a second target confidence of the pixel; and combining the far-distance disparity map and a disparity map formed by the new disparity value on a same unit into an index map, combining the first confidence and the first target confidence into a confidence map, optimizing the index and confidence maps to obtain a final disparity map, which is converted to a final depth map.