Apparatus and method for detecting obstacle through stereovision
According to an apparatus and method for detecting an obstacle through stereovision, an image capturing module comprises a plurality of cameras and is used for capturing a plurality of images; an image processing module edge-detecting the image to generate a plurality of edge objects and object information corresponding to each edge object; an object detection module matches a focus and a horizontal spacing interval of the camera according to the object information to generate a relative object distance corresponding to each edge object; a group module compares the relative object distance with a threshold distance and groups the edge objects with the relative object distance smaller than the threshold distance to be an obstacle and obtains a relative obstacle distance corresponding to the obstacle.
1. An apparatus for detecting an obstacle through stereovision, comprising:
an image capturing module, comprising a plurality of cameras and used for capturing a plurality of images;
an image processing module, edge-detecting the image and generating a plurality of edge objects and object information corresponding to each edge object;
an object detection module, matching a focus (f) and a horizontal spacing interval (B) of the two cameras in the plurality of cameras and generating a relative object distance corresponding to each edge object; and
a group module, comparing the relative object distance with a threshold distance, grouping the edge objects with the relative object distance smaller than the threshold distance to be an obstacle and obtaining a relative obstacle distance corresponding to the obstacle.
2. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the object information is selected from one of a group consisting of an object coordinate, image dot number, object length and object width of each edge object.
3. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the image processing module comprises a sampling unit used for lowering resolution of the image by means of decimation.
4. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the image processing module comprises a smoothing unit used for eliminating noise of the image by means of low-pass filtering.
5. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the image processing module comprises:
a horizontal edge detection unit used for detecting the image to generate horizontal line segments by means of an edge detection operator; and
a horizontal connection unit used for connecting the horizontal line segments to be a horizontal bounding box.
6. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the image processing module comprises:
a vertical edge detection unit used for detecting the image to generate vertical line segments by means of an edge detection operator; and
a vertical connection unit used for connecting the vertical line segments to be a vertical bounding box.
7. The apparatus for detecting an obstacle through stereovision according to claim 1 , wherein the object detection module comprises:
a comparison unit, comparing the object information to generate suspect objects; and
a searching unit, searching the image according to the suspect object to generate a disparity value (dx) of the edge object;
wherein, the relative object distance (Z) is obtained from a formula:
Z
=
B
×
f
dx
.
8. The apparatus for detecting an obstacle through stereovision according to claim 7 , wherein the cameras capturing the images is asynchronous in the image capturing module.
9. The apparatus for detecting an obstacle through stereovision according to claim 1 , further comprising a display module displaying the image and indicating the relative obstacle distance.
10. The apparatus for detecting an obstacle through stereovision according to claim 1 , further comprising an alarm unit used for setting a warning distance and comparing the warning distance with the relative obstacle distance to generate warning information.
11. A method for detecting an obstacle through stereovision, comprising the following steps:
(a) capturing a plurality of images through a plurality of cameras;
(b) edge-detecting the image to generate a plurality of edge objects and object information corresponding to each edge object;
(c) matching a focus (f) and a horizontal spacing interval (B) of the two cameras in the plurality of cameras according to the object information to generate a relative object distance corresponding to each edge object;
(d) comparing the relative object distance with a threshold distance; and
(e) grouping the edge objects with the relative object distance smaller than the threshold distance to be an obstacle and obtaining a relative obstacle distance corresponding to the obstacle.
12. The method for detecting an obstacle through stereovision according to claim 11 , wherein the object information is selected from one of a group consisting of an object coordinate, image dot number, object length and object width of each edge object.
13. The method for detecting an obstacle through stereovision according to claim 11 , wherein step (b) further comprises lowering resolution of the image by means of decimation.
14. The method for detecting an obstacle through stereovision according to claim 11 , wherein step (b) further comprises eliminating noise of the image by means of low-pass filtering.
15. The method for detecting an obstacle through stereovision according to claim 11 , wherein step (b) further comprises detecting the image by means of edge detection operator to generate horizontal line segments and connecting the horizontal line segments to be a horizontal bounding box.
16. The method for detecting an obstacle through stereovision according to claim 11 , wherein step (b) further comprises detecting the image by means of edge detection operator to generate vertical line segments and connecting the vertical line segments to be a vertical bounding box.
17. The method for detecting an obstacle through stereovision according to claim 11 , wherein step (c) further comprises the following steps:
comparing the object information to generate a suspect object; and
searching the image according to the suspect object to generate a disparity value (dx) of the edge object;
wherein, the relative object distance (Z) is obtained from a formula:
Z
=
B
×
f
dx
.
18. The method for detecting an obstacle through stereovision according to claim 17 , wherein the cameras capturing the images is asynchronous.
19. The method for detecting an obstacle through stereovision according to claim 11 , further comprising displaying the image and indicating the relative obstacle distance.
20. The method for detecting an obstacle through stereovision according to claim 11 , further comprising setting a warning distance and comparing the warning distance with the relative obstacle distance to generate warning information.