Binocular distance measuring method and device
A technology of binocular ranging and binocular camera, applied in the field of computer vision, can solve the problem of low accuracy of binocular ranging
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0070] see figure 2 , which is a flowchart of a binocular ranging method in an embodiment of the present invention, the method includes:
[0071] Step 101, perform feature extraction on the left image and the right image collected by the binocular camera to obtain a left feature map and a right feature map.
[0072] Step 102: Select one of the left feature map and the right feature map as a standard feature map, and use a block matching algorithm to perform correlation calculations on the left feature map and the right feature map to obtain the cost volume (ie cost volume) of the standard feature map.
[0073] Step 103: Perform normalization processing on the calculation costs of all difference values in the difference dimension of each pixel in the cost volume to obtain a confidence volume (ie, confidence volume).
[0074] Step 104, taking the maximum value of the confidence of all difference values on the difference dimension of each pixel in the confidence volume to o...
example 1
[0108] Take the left feature map as the standard feature map, take the right feature map as the reference feature map, set the sliding range to d, the horizontal moving direction to the horizontal left, and set the step size to 1 times the distance between two adjacent pixels; set the standard The pixels in the feature map are called standard pixels, the pixels in the reference feature map are called reference pixels, and the cost volume can be obtained by performing the following steps 1a to 3a for each standard pixel in the left feature map . Take the standard pixel point A as an example to perform the following steps 1a to 3a for description:
[0109] Step 1a. Slide to the left once in the right feature map based on the coordinates of the standard pixel point A to obtain the reference pixel point A1; slide to the left once again based on the previous reference pixel point A1 in the right feature map to obtain Reference pixel point A2; ..., slide to the left once based on t...
example 2
[0115] The right feature map is used as the standard feature map, the left feature map is used as the reference feature map, the sliding range is set to d, the horizontal moving direction is horizontal to the right, and the step size is set to 1 times the distance between two adjacent pixels; the standard The pixels in the feature map are called standard pixels, the pixels in the reference feature map are called reference pixels, and the cost volume can be obtained by performing the following steps 1b to 3b for each standard pixel in the right feature map . Take the standard pixel point B as an example to perform the following steps 1b to 3b for description:
[0116]Step 1b. Slide to the right once in the left feature map based on the coordinates of the standard pixel point B to obtain the reference pixel point B1; slide to the right once again based on the previous reference pixel point B1 in the left feature map to obtain Reference pixel point B2; ..., slide to the right on...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More