External environment recognition device for vehicle and vehicle system using same
An external environment recognition device for a vehicle and a vehicle system capable of assuring safety and reducing the processing load in a compatible manner are provided. An external environment recognition device 100 for a vehicle includes: first collision determination means 103 for computing a risk of collision of a host vehicle with a detected object in front of the host vehicle on the basis of information of a predicted course of the host vehicle and the detected object; and second collision determination means 104 for determining whether the detected object enters the predicted course from the outside of the predicted course or not. This device further includes object selection means 105 for selecting the detected object having a risk at least a first threshold and the detected object determined to enter the predicted course as selection candidate objects from among the detected objects, and for selecting the selection candidate object having a minimum relative distance to the host vehicle as a pedestrian determination request object from among the selected selection candidate objects. This device further includes pedestrian determination means 106 for determining whether the selected pedestrian determination request object is a pedestrian or not using image information.
1. An external environment recognition device for a vehicle that recognizes an external environment of the vehicle on the basis of detected object information of a plurality of detected objects in front of a host vehicle, image information of an image in front of the host vehicle, and host vehicle information of a detected status of the host vehicle, comprising:
predicted course setting means for setting a predicted course of the host vehicle on the basis of the host vehicle information;
first collision determination means for computing a risk of collision of each of the detected objects with the host vehicle on the basis of the predicted course set by the predicted course setting means and the detected object information;
second collision determination means for determining whether each of the detected objects enters the predicted course from outside of the predicted course or not on the basis of the predicted course and the detected object information;
object selection means for selecting the detected object having the risk computed by the first collision determination means at least a preset first threshold and the detected object determined by the second collision determination means to enter the predicted course as selection candidate objects among the detected objects, and for selecting the selection candidate object having a minimum relative distance to the host vehicle or a minimum predicted collision time as a pedestrian determination request object from among the selected selection candidate objects; and
pedestrian determination means for determining whether the pedestrian determination request object selected by the object selection means is a pedestrian or not using the image information;
wherein, in a case where the selection candidate object selected as the pedestrian determination request object in a last process is included in the selected selection candidate objects, the object selection means identifies the selected selection candidate object as a reselection candidate object, and determines whether to omit the reselection candidate object from a selection candidate of the pedestrian determination request object or not on the basis of the predicted course and the detected object information.
2. The external environment recognition device for a vehicle according to claim 1 , wherein the pedestrian determination means sets an image processing region of the image information according to a result of determination by the first collision determination means on the pedestrian determination request object and a result of determination by the second collision determination means.
3. The external environment recognition device for a vehicle according to claim 1 , wherein the pedestrian determination means, in a case where determining the pedestrian determination request object as a pedestrian, calculates camera information of the pedestrian determination request object on the basis of the image information.
4. The external environment recognition device for a vehicle according to claim 3 , further comprising fusion means for integrating the detected object information and the camera information and generating fusion information of the pedestrian determination request object determined as the pedestrian by the pedestrian determination means.
5. The external environment recognition device for a vehicle according to claim 4 , further comprising integrated collision determination means for computing a risk of collision of the host vehicle with the pedestrian determination request object according to the fusion information generated by the fusion means.
6. A vehicle system including the external environment recognition device for a vehicle according to claim 5 , comprising
control means for performing control for avoiding collision of the host vehicle in a case where the risk computed by the integrated collision determination means is less than a preset threshold.
7. The vehicle system according to claim 6 , wherein the control means computes a predicted collision time at which the host vehicle is to collide with the pedestrian determination request object on the basis of the fusion information generated by the fusion means, compares the computed predicted collision time with a preset breaking threshold, and, if the predicted collision time is less than or equal to the breaking threshold, performs a brake control on the vehicle.
8. The vehicle system according to claim 7 , wherein, if the predicted collision time is larger than the breaking threshold, the control means compares the predicted collision time with a preset warning threshold, and, if the predicted collision time is less than or equal to the warning threshold, performs a control of an alarm to the driver.