Systems and methods for object detection using stereovision information
Systems and methods for object detection. The methods comprise, by a computing device: obtaining a plurality of intensity values denoting at least a difference in a first location of at least one object in a first image and a second location of the at least one object in a second image; converting the intensity values to 3D position values; inputting the 3D position values into a classifier algorithm to obtain classifications for data points of a 3D point cloud (each of the classifications comprising a foreground classification or a background classification); and using the classifications to detect at least one object which is located in a foreground or a background.
1. A method for object detection, comprising:
obtaining, by a computing device, a plurality of intensity values denoting at least a difference in a first location of at least one object in a first image and a second location of the at least one object in a second image;
converting, by the computing device, the intensity values to 3D position values;
inputting, by the computing device, the 3D position values and the intensity values into a classifier algorithm to obtain classifications for data points of a 3D point cloud, each of the classifications comprising a foreground classification or a background classification; and
using, by the computing device, the classifications to detect at least one object which is located in a foreground or a background.
2. The method according to claim 1 , further comprising using the classifications to control operations of an autonomous vehicle.
3. The method according to claim 1 , wherein the plurality of intensity values is obtained from a stereo disparity image that is generated based on the first and second images captured concurrently by stereo cameras.
4. The method according to claim 1 , wherein each of the intensity values denotes a position difference between coordinates of a pixel in the first image and coordinates of a corresponding pixel in the second image.
5. The method according to claim 1 , wherein each of the intensity values denotes a distance between a pixel of the first or second image and a location of a camera that captured the first or second image.
6. The method according to claim 1 , wherein the converting comprises projecting each of said intensity values into an xyx-position in a 3D road map.
7. The method according to claim 6 , wherein the projecting is based on a time of image capture, a location of a vehicle at the time of image capture, positions of cameras relative to the vehicle, pixel pointing directions, and distances of the pixels from the cameras.
8. The method according to claim 1 , wherein the classifier algorithm comprises a hierarchical decision tree classifier.
9. The method according to claim 1 , further comprising estimating a height of each said 3D position value from ground using a 3D road map and inputting the estimated heights into the classifier algorithm along with the 3D position values.
10. The method according to claim 1 , wherein each of the intensity values comprises a scalar value.
11. A system, comprising:
a processor; and
a non-transitory computer-readable storage medium comprising programming instructions that are configured to cause the processor to implement a method for object detection, wherein the programming instructions comprise instructions to:
obtain a plurality of intensity values denoting at least a difference in a first location of at least one object in a first image and a second location of the at least one object in a second image;
convert the intensity values to 3D position values;
input the 3D position values and the intensity values into a classifier algorithm to obtain classifications for data points of a 3D point cloud, each of the classifications comprising a foreground classification or a background classification; and
use the classifications to detect at least one object which is located in a foreground or a background.
12. The system according to claim 11 , wherein the programming instructions comprise instructions to use the classifications to control operations of an autonomous vehicle.
13. The system according to claim 11 , wherein the plurality of intensity values is obtained from a stereo disparity image that is generated based on the first and second images captured concurrently by stereo cameras.
14. The system according to claim 11 , wherein each of the intensity values denotes a position difference between coordinates of a pixel in the first image and coordinates of a corresponding pixel in the second image.
15. The system according to claim 11 , wherein each of the intensity values denotes a distance between a pixel of the first or second image and a location of a camera that captured the first or second image.
16. The system according to claim 11 , wherein the converting comprises projecting each of said intensity values into an xyx-position in a 3D road map.
17. The system according to claim 16 , wherein the projecting is based on a time of image capture, a location of a vehicle at the time of image capture, positions of cameras relative to the vehicle, pixel pointing directions, and distances of the pixels from the cameras.
18. A non-transitory computer-readable medium that stores instructions configured to, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
obtaining a plurality of intensity values denoting at least a difference in a first location of at least one object in a first image and a second location of the at least one object in a second image;
converting the intensity values to 3D position values;
inputting the 3D position values and the intensity values into a classifier algorithm to obtain classifications for data points of a 3D point cloud, each of the classifications comprising a foreground classification or a background classification; and
using the classifications to detect at least one object which is located in a foreground or a background.