Passive stereo depth sensing
AssigneeCompound Eye, Inc.
InventorJason Devitt
Life of this patent
Estimating the term…
Abstract
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.
Claims20
A method of generating a depth map representing distances between an image acquisition system and objects in a scene captured by the image acquisition system, the method comprising:
- receiving a stereoscopic image pair comprising a first image and a second image, the first image and the second image captured by the image acquisition system at a substantially same time, the first image capturing a view of a scene from a first perspective and the second image capturing a view of the scene from a second perspective;
- generating binary hashes for pixels in the first and second images, wherein generating a binary hash for a pixel in an image comprises performing a random walk on a pixel patch of said image, wherein the pixel patch surrounds the pixel;
- matching pixels in the first and second images that depict same points in the scene responsive to cost values representing differences between binary hashes for the pixels;
- and
- generating the depth map representing distances between the image acquisition system and the objects in the scene responsive to the matching pixels, the depth map generated from a disparity map, the disparity map comprising a plurality of disparity vectors representing coordinate differences between matched pixels in the first and second images.
A non-transitory computer-readable storage medium storing executable computer program instructions for generating a depth map representing distances between an image acquisition system and objects in a scene captured by the image acquisition system, the instructions when executed by at least one computer processor causing the computer processor to perform steps comprising:
- receiving a stereoscopic image pair comprising a first image and a second image, the first image and the second image captured by the image acquisition system at a substantially same time, the first image capturing a view of a scene from a first perspective and the second image capturing a view of the scene from a second perspective;
- generating binary hashes for pixels in the first and second images, wherein generating a binary hash for a pixel in an image comprises performing a random walk on a pixel patch of said image, wherein the pixel patch surrounds the pixel;
- matching pixels in the first and second images that depict same points in the scene responsive to cost values representing differences between binary hashes for the pixels;
- and
- generating the depth map representing distances between the image acquisition system and the objects in the scene responsive to the matching pixels, the depth map generated from a disparity map, the disparity map comprising a plurality of disparity vectors representing coordinate differences between matched pixels in the first and second images.
A system for generating a depth map representing distances between an image acquisition system and objects in a scene captured by the image acquisition system, the system comprising:
- at least one computer processor;
- and
- a non-transitory computer-readable storage medium storing executable computer program instructions, the instructions when executed by the computer processor causing the computer processor to perform steps comprising:
- receiving a stereoscopic image pair comprising a first image and a second image, the first image and the second image captured by the image acquisition system at a substantially same time, the first image capturing a view of a scene from a first perspective and the second image capturing a view of the scene from a second perspective,
- generating binary hashes for pixels in the first and second images, wherein generating a binary hash for a pixel in an image comprises performing a random walk on a pixel patch of said image, wherein the pixel patch surrounds the pixel,
- matching pixels in the first and second images that depict same points in the scene responsive to cost values representing differences between binary hashes for the pixels, and
- generating the depth map representing distances between the image acquisition system and the objects in the scene responsive to the matching pixels, the depth map generated from a disparity map, the disparity map comprising a plurality of disparity vectors representing coordinate differences between matched pixels in the first and second images.
Dependencies are inferred from the claim text.