Device and method for searching parking space

1. A device for searching a parking space, the device comprising:

at least one space detection sensor mounted on a vehicle; and

a controller configured to:

analyze sensed information obtained through the at least one space detection sensor and to recognize space and object information within a parking lot;

predict a distribution of available parking spaces in the parking lot based on the recognized space and object information; and

determine an optimal available parking space based on the distribution of the available parking spaces and characteristics of a driver,

wherein the controller is further configured to:

analyze the sensed information obtained through the at least one space detection sensor of the vehicle and generate a parking lot partial map including a parking line, locations of parked vehicles, and a location of a host vehicle,

select a recognized road in the parking lot partial map as a travelable space;

reflect an available parking space among the available parking spaces recognized while the vehicle travels to the travelable space in the parking lot partial map,

assign a weight accordingly when the available parking spaces are successively recognized, and reflect the weight on the parking lot partial map, and

select a parking space node on a travelable space closest to the available parking space.

2. The device of claim 1 , wherein the at least one space detection sensor includes at least one of a camera, a LiDar, or an ultrasonic sensor.

3. The device of claim 1 , wherein the controller is further configured to:

select an intersection or a turning point in the parking lot partial map as a fixed node;

select a boarding and alighting node at a driver boarding and alighting location in the parking lot partial map; and

select a cart node at a cart storage in the parking lot partial map.

4. The device of claim 3 , wherein the controller is further configured to determine the optimal available parking space in consideration of a location of the available parking space by the parking space node, the driver boarding and alighting location by the boarding and alighting node, or a location of the cart storage by the cart node.

5. The device of claim 4 , wherein the controller is further configured to determine an available parking space among the available parking spaces with the lowest cost function after setting a location of the parking space node, a location of the boarding and alighting node, and a location of the cart node as model inputs for convolutional neural network calculation as an optimal parking space.

6. A method for searching a parking space, the method comprising:

analyzing sensed information obtained through a space detection sensor including at least one of a camera, a LiDar, or an ultrasonic sensor installed in a vehicle and recognizing space and object information within a parking lot;

predicting a distribution of available parking spaces in the parking lot based on the recognized space and object information; and

determining an optimal available parking space based on the distribution of the available parking spaces and characteristics of a driver,

wherein the analyzing of sensed information obtained through the space detection sensor including at least one of the camera, the LiDar, or the ultrasonic sensor installed in the vehicle and recognizing space and object information within the parking lot includes:

analyzing the sensed information obtained through the space detection sensor of the vehicle to generate a parking lot partial map including a parking line, locations of parked vehicles, and a location of a host vehicle, and

wherein the predicting of the distribution of the available parking spaces in the parking lot based on the recognized space and object information includes:

selecting a recognized road in the parking lot partial map as a travelable space;

reflecting an available parking space among the available parking spaces recognized while the vehicle travels to the travelable space in the parking lot partial map;

assigning a weight accordingly when the available parking spaces are successively recognized, and reflecting the weight on the parking lot partial map; and

selecting a parking space node on a travelable space closest to the available parking space.

7. The method of claim 6 , wherein the predicting of the distribution of the available parking spaces in the parking lot based on the recognized space and object information further includes:

selecting an intersection or a turning point in the parking lot partial map as a fixed node;

selecting a boarding and alighting node at a driver boarding and alighting location in the parking lot partial map; and

selecting a cart node at a cart storage in the parking lot partial map.

8. The method of claim 7 , wherein the determining of the optimal available parking space based on the distribution of the available parking spaces and the characteristics of the driver includes:

determining the optimal available parking space in consideration of a location of the available parking space by the parking space node, the driver boarding and alighting location by the boarding and alighting node, or a location of the cart storage by the cart node.

9. The method of claim 8 , wherein the determining of the optimal available parking space based on the distribution of the available parking spaces and the characteristics of the driver includes:

determining an available parking space among the available parking spaces with the lowest cost function after setting a location of the parking space node, a location of the boarding and alighting node, and a location of the cart node as model inputs for convolutional neural network calculation as an optimal parking space.