Passive Stereo Depth Sensing via Random Walk Hashing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional passive depth sensing techniques are computationally intensive, making them slow and limiting their adoption in applications such as autonomous vehicles and augmented reality, where rapid depth information generation is required.
Innovation Solution
A method that generates a disparity map by receiving a stereoscopic image pair, performing random walks to create binary hashes for pixels, matching pixels based on cost values, and generating disparity vectors to represent distances between the image acquisition system and objects in the scene, thereby reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional passive depth sensing techniques are used, then depth information can be obtained, but computational requirements are high and processing time is long
Solution Approach 1:
The patent segments the image processing task by dividing images into superpixels (groups of pixels) and processing them collectively rather than individually. This segmentation reduces the computational burden while maintaining depth sensing accuracy, as similar pixels are processed together through random walk algorithms.
Solution Approach 2:
The patent replaces traditional mechanical/optical depth sensing methods with computational algorithms. Specifically, it substitutes conventional correlation-based matching with random walk algorithms that traverse pixel intensity landscapes, achieving faster and more accurate depth calculation without physical depth sensors.
2Measurement precision
If conventional passive depth sensing techniques are used, then depth sensing is achieved, but computational complexity is high limiting real-time application
Solution Approach 1:
The patent changes the computational parameters by using random walk algorithms that iterate through pixel intensities with probabilistic transitions. This approach transforms the complex optimization problem into a series of simpler stochastic steps, reducing overall computational complexity while maintaining depth sensing precision.
Solution Approach 2:
The patent performs preliminary actions by pre-processing images into superpixels and pre-computing intensity relationships before actual depth calculation. This preliminary organization of data structures and intensity mappings reduces the complexity of subsequent depth computation operations.
Data Source
AI summary
A depth sensing system receives an image pair showing objects in a scene. The system generates binary hashes for pixels in the image pair by performing a random walk. The system matches pixels in the first image to pixels in the second image that depict the same point in the scene by generating cost values representing differences between the binary hashes for pairs of pixels in the images. The system generates a disparity map containing disparity vectors representing coordinate differences between matched pixels in the first and second images. The system generates and outputs a depth map based on the disparity map. The depth map represents the distances between an image acquisition system that acquired the image pair and the objects in the scene.