Object detecting device

1. An object detecting system comprising:

two cameras configured to capture images in front of a vehicle; and

a processor configured to:

compare individual images of the two cameras and calculate a disparity for each pixel;

set a boundary between a near region close to the vehicle and a far region distant from the vehicle, in a single image of one of the two cameras;

detect an object of the near region, on the basis of the disparity; and

detect an object of the far region, on the basis of the single image,

wherein the processor configured to set the boundary further comprises the processor configured to:

output a V-Disparity image obtained by projecting, onto a virtual plane where a first axis shows the disparity and a second axis shows a coordinate position of a longitudinal direction of each pixel of the single image, data of the disparity and the coordinate position corresponding to each pixel of the single image.

2. The object detecting system according to claim 1 , wherein the processor is further configured to:

erase a detection result of a long distance in detection results of the objects displayed on the single image and overlapping each other.

3. The object detecting system according to claim 1 , wherein the processor is further configured to:

extract, using a disparity image in which a disparity calculated for each pixel of the near region of the single image is stored for each pixel, a group of the distance data in which distance data to be the disparity of each pixel of the disparity image is continuous in a depth direction and a transverse direction, from the disparity image, and detects the object of the near region.

4. An object detecting device for detecting objects in front of a vehicle, the object detecting device comprising:

a processor configured to:

compare individual images of two cameras and calculate a disparity for each pixel, wherein the individual images are captured by the two cameras;

set a boundary between a near region close to the vehicle and a far region distant from the vehicle, in a single image of one of the two cameras;

detect an object of the near region, on the basis of the disparity; and

detect an object of the far region, on the basis of the single image,

wherein the processor is further configured to:

calculate a distance to a road surface in front of the vehicle, on the basis of the disparity, and detect a boundary pixel position between a region where reliability of the distance is high and a region where the reliability of the distance is low; and

set the boundary on the basis of the boundary pixel position, and

wherein the processor is further configured to:

output a V-Disparity image obtained by projecting, onto a virtual plane where a transverse axis shows the disparity and a longitudinal axis shows a coordinate position of a longitudinal direction of each pixel of the single image, data of the disparity and the coordinate position corresponding to each pixel of the single image;

detect a most dominant straight line in the V-Disparity image; and

detect the boundary pixel position, on the basis of a deviation of the straight line in the V-Disparity image and a position of the data.

5. The object detecting device according to claim 4 , wherein the processor is further configured to:

set a search range to the far region on the single image, scan a template in the search range, and detect the object of the far region by pattern matching.

6. The object detecting device according to claim 5 , wherein the processor is further configured to:

set using an estimation error between a coordinate position of a longitudinal direction of a pixel at the boundary of the single image and the boundary pixel position of the V-Disparity image, one end of a longitudinal direction of the search range to a position to be a sum of the boundary pixel position and the estimation error.

7. The object detecting device according to claim 6 , wherein the processor is further configured to:

set the other end of the longitudinal direction of the search range to a position obtained by subtracting an image longitudinal width of an object of a maximum height assumed at the boundary of the single image and the estimation error from the boundary pixel position, using the image longitudinal width.

8. The object detecting device according to claim 5 , wherein the processor is further configured to:

set a road shape in front of the vehicle using the single image,

wherein the processor sets positions of left and right ends of the search range to positions of left and right ends of an image region of the road shape.

9. The object detecting device according to claim 5 , wherein the processor is further configured to:

scan the search range repetitively using a plurality of templates having different sizes.

10. The object detecting device according to claim 4 , wherein the processor is further configured to:

erase a detection result of a long distance in detection results of the objects displayed on the single image and overlapping each other.

11. The object detecting device according to claim 4 , wherein the processor is further configured to:

extract, using a disparity image in which a disparity calculated for each pixel of the near region of the single image is stored for each pixel, a group of the distance data in which distance data to be the disparity of each pixel of the disparity image is continuous in a depth direction and a transverse direction, from the disparity image, and detects the object of the near region.