Back to results
US 11,308,638 B2Granted
Depth estimation method and apparatus, electronic device, program, and medium
AssigneeShenzhen Sensetime Technology Co., Ltd.
InventorsYue Luo, Sijie Ren
Life of this patent
Estimating the term…
Abstract
A depth estimation method includes: taking a single image as a first image in binocular images, and obtaining a second image in the binocular images based on the first image via a first neural network; and obtaining depth information corresponding to the first image via a second neural network by performing binocular stereo matching on the first image and the second image.
Claims20
A depth estimation method, comprising:
- taking a single image as a first image in binocular images, and obtaining a second image in the binocular images based on the first image via a first neural network;
- and
- obtaining depth information corresponding to the first image via a second neural network by performing binocular stereo matching on the first image and the second image,
- wherein the obtaining a second image in the binocular images based on the first image via a first neural network comprises:
- processing the first image via the first neural network to output disparity probability maps of N channels, wherein the disparity probability map of each channel represents a probability of a pixel on the first image being shifted by i disparities in a first horizontal direction, wherein i=0, 1, . . . , N−1, and N is an integer greater than 1;
- shifting the first image by i pixels in the first horizontal direction respectively according to the disparity probability maps of the N channels to obtain N shifted images;
- dot-multiplying each of the N shifted images by the disparity probability map of the corresponding channel to obtain N dot multiplication results;
- and
- superimposing the N dot multiplication results based on pixels to obtain the second image.
6 more elements
A depth estimation apparatus, comprising:
- a memory storing processor-executable instructions;
- and
- a processor arranged to execute the stored processor-executable instructions to perform operations of:
- taking a single image as a first image in binocular images, and obtaining a second image in the binocular images based on the first image via a first neural network;
- and
- obtaining depth information corresponding to the first image via a second neural network by performing binocular stereo matching on the first image and the second image,
- wherein the obtaining a second image in the binocular images based on the first image via a first neural network comprises:
- processing the first image via the first neural network to output disparity probability maps of N channels, wherein the disparity probability map of each channel represents a probability of a pixel on the first image being shifted by i disparities in a first horizontal direction, wherein i=0, 1, . . . , N−1, and N is an integer greater than 1;
- shifting the first image by i pixels in the first horizontal direction respectively according to the disparity probability maps of the N channels to obtain N shifted images;
- dot-multiplying each of the N shifted images by the disparity probability map of the corresponding channel to obtain N dot multiplication results;
- and
- superimposing the N dot multiplication results based on pixels to obtain the second image.
9 more elements
A non-transitory computer-readable storage medium having stored thereon computer-readable instructions that, when executed by a processor, cause the processor to perform a depth estimation method, the method comprising:
- taking a single image as a first image in binocular images, and obtaining a second image in the binocular images based on the first image via a first neural network;
- and
- obtaining depth information corresponding to the first image via a second neural network by performing binocular stereo matching on the first image and the second image,
- wherein the obtaining a second image in the binocular images based on the first image via a first neural network comprises:
- processing the first image via the first neural network to output disparity probability maps of N channels, wherein the disparity probability map of each channel represents a probability of a pixel on the first image being shifted by i disparities in a first horizontal direction, wherein i=0, 1, . . . , N−1, and N is an integer greater than 1;
- shifting the first image by i pixels in the first horizontal direction respectively according to the disparity probability maps of the N channels to obtain N shifted images;
- dot-multiplying each of the N shifted images by the disparity probability map of the corresponding channel to obtain N dot multiplication results;
- and
- superimposing the N dot multiplication results based on pixels to obtain the second image.
6 more elements
Dependencies are inferred from the claim text.