Image processing device and image processing method
An image processing device includes a sideline detector configured to detect sidelines of boundary lines of parking sections next to each other in a vehicle width direction by an edge detection process to an image, each of the sidelines being a pair of a rising edge and a falling edge, and to acquire an end point of each of the sidelines at which the pair is not detected, a rear-end edge identifying part configured to identify a rear-end edge of the parking section based on an end point away from the vehicle among the end points acquired by the sideline detector, and a parking frame setting portion configured to identify the parking section based on the two sidelines detected by the sideline detector and the rear-end edge identified by the rear-end edge identifying part, and to set a parking frame based on the parking section.
1. An image processing device for setting a parking frame of a parking target corresponding to a parking section for a vehicle based on an image captured by a camera mounted on the vehicle, the image processing device comprising:
a processor; and
a non-transitory computer-readable medium having stored thereon executable instructions that, when executed by the processor, cause the image processing device to function as:
a marker detector configured to: (i) detect sidelines that are boundary lines of parking sections next to each other in a vehicle width direction by an edge detection process to the image, (ii) acquire an end point of each of the sidelines at which the sideline is not detected,
and (iii) detect a rear-end edge of the parking section based on an end point away from the vehicle among the end points; and
a parking frame setting portion configured to set the parking frame based on the sidelines and the rear-end edge detected by the marker detector.
2. The device according to claim 1 , wherein the marker detector is configured to determine whether or not a virtual straight line connecting the two end points of the two sidelines is orthogonal to the two sidelines, and to identify the virtual straight line as the rear-end edge when determining that the virtual straight line is orthogonal to the two sidelines.
3. The device according to claim 1 , wherein the marker detector is configured to determine whether or not a virtual straight line connecting the two end points of the two sidelines is orthogonal to the two sidelines, to search a straight line orthogonal to the two sidelines around the two end points of the two sidelines when determining that the virtual straight line is not orthogonal to the two sidelines, and to identify the straight line orthogonal to the two sidelines as the rear-end edge when detecting the straight line orthogonal to the two sidelines.
4. The device according to claim 1 , wherein the marker detector is configured to search a straight line along a virtual straight line connecting the two end points of the two sidelines around the two end points of the two sidelines, and to identify the straight line as the rear-end edge when detecting the straight line along the virtual straight line.
5. The device according to claim 1 , wherein the marker detector is configured to search an extended line along the sideline ahead of the end point of a first of the two sidelines, to search a straight line that is along a virtual straight line connecting an end point of the extended line and the end point of a second of the two sidelines and is orthogonal to the sidelines, and to identify the straight line as the rear-end edge when detecting the straight line that is along the virtual straight line and is orthogonal to the sidelines.
6. An image processing method for setting a parking frame of a parking target corresponding to a parking section for a vehicle based on an image captured by a camera mounted on the vehicle, the image processing method comprising:
detecting sidelines that are boundary lines of parking sections next to each other in a vehicle width direction by an edge detection process to the image, acquiring an end point of each of the sidelines at which the sideline is not detected,
and detecting a rear-end edge of the parking section based on an end point away from the vehicle among the end points; and
setting the parking frame based on the sidelines and the rear-end edge which have been detected.
7. The method according to claim 6 , further comprising determining whether or not a virtual straight line connecting the two end points of the two sidelines is orthogonal to the two sidelines, and identifying the virtual straight line as the rear-end edge when it is determined that the virtual straight line is orthogonal to the two sidelines.
8. The method according to claim 6 , further comprising determining whether or not a virtual straight line connecting the two end points of the two sidelines is orthogonal to the two sidelines, searching a straight line orthogonal to the two sidelines around the two end points of the two sidelines when it is determined that the virtual straight line is not orthogonal to the two sidelines, and identifying the straight line orthogonal to the two sidelines as the rear-end edge when the straight line orthogonal to the two sidelines is detected.
9. The method according to claim 6 , further comprising searching a straight line along a virtual straight line connecting the two end points of the two sidelines around the two end points of the two sidelines, and identifying the straight line as the rear-end edge when the straight line along the virtual straight line is detected.
10. The method according to claim 6 , further comprising searching an extended line along the sideline ahead of the end point of a first of the two sidelines, searching a straight line that is along a virtual straight line connecting an end point of the extended line and the end point of a second of the two sidelines and is orthogonal to the sidelines, and identifying the straight line as the rear-end edge when the straight line that is along the virtual straight line and is orthogonal to the sidelines is detected.