Image processor and image processing method

1. An image processor comprising:

a control portion configured to detect a marker drawn on a road surface from an image captured by an imaging device that captures the road surface around a vehicle,

wherein the control portion connects a plurality of broken markers to create a single marker when the detected marker is broken into plural, wherein

the control portion comprises:

an edge detection portion configured to scan the captured image in a predetermined direction to detect a plurality of edges;

an edge connection portion configured to detect a closest edge which is closest to a subject edge selected from the plurality of edges and that connects the subject edge and the closest edge to create a single edge when a distance between endpoints of the subject edge and the closest edge is predetermined interval or shorter; and

a marker detection portion configured to detect the marker based on the plurality of edges,

wherein the edge connection portion is configured to, when each of a distance from a line, which is formed by extending the subject edge, to a start point of the closest edge and a distance from the line, which is formed by extending the subject edge, to an end point of the closest edge is a threshold or below, to connect the subject edge and the closest edge.

2. The image processor according to claim 1 , wherein the edge detection portion scans an area between the endpoints of the subject edge and the closest edge in a direction in which the subject edge extends and determines whether or not information related to the marker is present between the subject edge and the closest edge.

3. The image processor according to claim 1 , wherein the edge connection portion counts a number of edges present within a predetermined range from the subject edge as an edge point, and the edge connection portion does not connect the subject edge with any of the plurality of edges when the counted edge point is a threshold number or greater.

4. The image processor according to claim 1 , wherein

the edge connection portion assigns the endpoint of the subject edge, which is opposite to the endpoint of the subject edge close to the closest edge, as the coordinate of a start point of the single edge, and

when a distance between a straight line formed by extending the subject edge and an X-coordinate of the endpoint of the closest edge, which is opposite to the endpoint of the closest edge close to the subject edge, is a threshold value or shorter, the edge connection portion assigns a finish point of the closest edge as a coordinate of a finish point of the single edge,

whereas when the distance between the straight line and the X-coordinate of the endpoint of the closest edge, which is opposite to the endpoint of the closest edge close to the subject edge, exceeds the threshold value, the edge connection portion assigns an intersection of the straight line and a line extending perpendicularly from the X-coordinate of the endpoint of the closest edge, which is opposite to the endpoint of the closest edge close to the subject edge, to the straight line as the coordinate of the finish point of the single edge.

5. The image processor according to claim 1 ,

wherein the control portion comprises:

a marker detection portion configured to detect the marker from the captured image; and

a correction portion, wherein

the correction portion determines whether or not a further marker is detected on an extension line of the marker from a position separated from the marker by a preset distance, and

the correction portion connects the marker and the further marker to create the single marker when the further marker is detected.

6. The image processor according to claim 5 , wherein the correction portion calculates a brightness difference between adjacent pixels from an endpoint of the marker detected by the marker detection portion along an extending direction of the maker and determines that the marker is broken when an absolute value of the calculated brightness difference is a preset value or greater, and

the correction portion determines whether or not the further marker is detected on the extension line of the marker from the position separated from the marker by the preset distance when it is determined that the maker is broken.

7. The image processor according to claim 6 , wherein the preset value, which is used by the correction portion to determine whether or not the marker is broken, is smaller than a prescribed value, which is used by the marker detection portion to detect the marker.

8. The image processor according to claim 1 , wherein the marker is a parking area line that defines a parking space, and

the image processor further comprises a parking frame setting portion that sets a parking frame based on the detected marker.

9. The image processor according to claim 1 , wherein the marker is a lane borderline that divides a traveling lane, and

the image processor further comprises a traveling lane detection portion that detects the traveling lane based on the detected marker.

10. An image processing method, comprising:

detecting a marker drawn on a road surface from an image captured by an imaging device that captures the road surface around a vehicle; and

image-processing to connect a plurality of broken markers to create a single marker when the detected marker is broken into plural, wherein the image-processing comprises:

detecting a plurality of edges by scanning the captured image in a predetermined direction;

detecting a closest edge which is closest to a subject edge selected from the plurality of edges;

connecting the subject edge and the closest edge to create a single edge when a distance between endpoints of the subject edge and the closest edge is predetermined interval or shorter; and

detecting the marker based on the plurality of edges,

wherein, when each of a distance from a line, which is formed by extending the subject edge, to a start point of the closest edge and a distance from the line, which is formed by extending the subject edge, to an end point of the closest edge is a threshold or below, connecting the subject edge and the closest edge.

11. The image processing method according to claim 10 , wherein the image-processing comprises:

detecting the marker from the captured image;

determining whether or not a further marker is detected on an extension line of the marker from a position separated from the marker by a preset distance; and

connecting the marker and the further marker to create the single marker when the further marker is detected.