Method and device for predicting path of surrounding object

1 . A method of predicting a path of an object in surroundings of a vehicle, the method comprising:

recognizing, by at least one sensor of the vehicle, the object;

generating, by an electronic control device including at least one processor, movement path data associated with the object by tracking, in response to the object being recognized, a movement path of the object during a first time interval and classifying the object into a first type corresponding to a vehicle or a second type corresponding to a pedestrian;

extracting, by the electronic control device, a road map of a road on which the vehicle is positioned; and

generating, by the electronic control device and based on the generated movement path data and the road map, prediction path data including at least one prediction path associated with the object during a second time interval,

wherein the generating of the prediction path data comprises:

when the object is classified into the first type, generating, by the electronic control device, the prediction path data using a second artificial neural network model trained to generate the prediction path data for the vehicle; and

when the object is classified into the second type, generating, by the electronic control device, the prediction path data using a first artificial neural network model trained to generate the prediction path data for the pedestrian,

wherein the generating of the prediction path data comprises:

determining, by the electronic control device, a first prediction path and a second prediction path through which the object is expected to move by utilizing the movement path data and the road map; and

determining, by the electronic control device, a first probability that the object is expected to move through the determined first prediction path and a second probability that the object is expected to move through the determined second prediction path,

wherein the generating of the prediction path data further comprises

generating, by the electronic control device, the prediction path data by providing the movement path data to the first artificial neural network model or the second artificial neural network model, and

wherein the generating of the prediction path data further comprises:

when the object is classified into the first type, generating, by the electronic control device, the prediction path data by providing the movement path data and the road map to the second artificial neural network model; and

when the object is classified into the second type, generating, by the electronic control device, the prediction path data by providing the movement path data to the first artificial neural network model.

2 . The method of claim 1 , further comprising:

determining, by the electronic control device, a driving trajectory of the vehicle during the second time interval;

calculating, by the electronic control device and in response to the first probability being equal to or more than a first threshold, a collision risk of the vehicle and the object colliding based on the determined driving trajectory and the first prediction path; and

changing, by the electronic control device and in response to the calculated collision risk being equal to or more than a second threshold, the determined driving trajectory.

3 . The method of claim 2 , further comprising:

calculating, by the electronic control device, prediction speed data of the object associated with the first prediction path based on the generated prediction path data,

wherein the calculating of the collision risk comprises

calculating, by the electronic control device and in response to the prediction speed data, the collision risk based on a probability that the vehicle and the object are expected to collide with each other in response to the object moving through the first prediction path.

4 . The method of claim 1 , further comprising:

transferring, by the electronic control device and in response to a difference between the first probability and the second probability being equal to or less than a third threshold, a control right of the vehicle to a driver of the vehicle.

5 . The method of claim 1 , wherein the extracting of the road map comprises

determining, by the electronic control device, a current position of the vehicle based on the at least one sensor; and

extracting, by the electronic control device, the road map within a predetermined range at the determined current position of the vehicle.

6 . The method of claim 1 , further comprising:

determining, by the electronic control device, a driving trajectory of the vehicle during the second time interval;

calculating, by the electronic control device, a collision risk of the vehicle and the object colliding based on the determined driving trajectory and the at least one prediction path; and

changing, by the electronic control device, the determined driving trajectory in response to the calculated collision risk being equal to or more than a threshold.

7 . A vehicle system comprising:

at least one sensor configured to recognize an object in surroundings of a vehicle;

an electronic control device configured to generate movement path data associated with the object by tracking, when the object is recognized, a movement path of the object during a first time interval, classifying the object into a first type corresponding to a vehicle or a second type corresponding to a pedestrian, and generating prediction path data including at least one prediction path associated with the object during a second time interval based on the generated movement path data to generate a driving trajectory to prevent the vehicle and the object from colliding with each other; and

a driving unit configured to drive the vehicle based on the generated driving trajectory when the driving trajectory of the vehicle is generated by the electronic control device,

wherein the electronic control device is configured to:

when the object is classified into the first type, generate the prediction path data using a second artificial neural network model trained to generate the prediction path data for a vehicle; and

when the object is classified into the second type, generate the prediction path data using a first artificial neural network model trained to generate the prediction path data for a pedestrian,

wherein the electronic control device is configured to:

extract a road map of a road on which the vehicle is positioned;

generate the prediction path data based on the movement path data and the road map;

determine a first prediction path and a second prediction path through which the object is expected to move by using the movement path data and the road map; and

determine a first probability that the object is expected to move through the determined first prediction path and a second probability that the object is expected to move through the determined second prediction path,

wherein the electronic control device is further configured to

generate the prediction path data by providing the movement path data to the first artificial neural network model or the second artificial neural network model, and

wherein the electronic control device is configured to:

generate the prediction path data by providing the movement path data and the road map to the second artificial neural network model when the object is classified into the first type; and

generate the prediction path data by providing the movement path data to the first artificial neural network model when the object is classified into the second type.

8 . The vehicle system of claim 7 , wherein the electronic control device is configured to

determine a driving trajectory of the vehicle during the second time interval,

calculate a collision risk of the vehicle and the object colliding based on the determined driving trajectory and the first prediction path when the first probability is equal to or more than a first threshold, and

change the determined driving trajectory when the calculated risk is equal to or more than a second threshold.

9 . The vehicle system of claim 8 , wherein the electronic control device is configured to

calculate prediction speed data of the object associated with the first prediction path based on the generated prediction path data, and

calculate, in response to the prediction speed data, the collision risk based on a probability that the vehicle and the object are expected to collide with each other when the object moves through the first prediction path.

10 . The vehicle system of claim 7 , wherein when a difference between the first probability and the second probability is equal to or less than a third threshold, the electronic control device is configured to transfer a control right of the vehicle to a driver of the vehicle.

11 . The vehicle system of claim 7 , wherein the electronic control device is configured to

determine a current position of the vehicle based on the at least one sensor, and

extract a road map within a predetermined range at the determined current position of the vehicle.

12 . The vehicle system of claim 7 , wherein the electronic control device is configured to

determine a driving trajectory of the vehicle during the second time interval,

calculate the collision risk of the vehicle and the object colliding based on the determined driving trajectory of the vehicle and the at least one prediction path, and

change the determined driving trajectory of the vehicle when the calculated collision risk is equal to or more than a threshold.