Systems and methods for object classification in autonomous vehicles
Systems and method are provided for controlling a vehicle. In one embodiment, an object classification method includes receiving sensor data associated with an object observed by a sensor system of an autonomous vehicle and determining, with a processor, a bounding curve associated with the sensor data. A plurality of bounding curve features are determined based on a set of convexities and concavities associated with the bounding curve. The object is classified by applying the plurality of bounding curve features to a machine learning model and receiving a classification output.
1. An object classification method comprising:
receiving lidar point cloud data associated with an object observed by a sensor system of an autonomous vehicle;
determining, with a processor, a bounding curve as an outline of the object in the lidar point cloud data and having a plurality of curve segments;
determining, with the processor, a feature vector for each of the plurality of curve segments based on convexities and concavities associated with the bounding curve; and
classifying, with the processor, the object by applying the feature vectors to a machine learning model and receiving a classification output that classifies the object for assisting in control of the autonomous vehicle.
2. The method of claim 1 , wherein the machine learning model is an artificial neural network model.
3. The method of claim 1 , further including transmitting the machine learning model to the autonomous vehicle over a communication network.
4. The method of claim 1 , wherein the feature vectors include gradients between adjacent convexities and concavities.
5. The method of claim 1 , wherein the feature vectors include the elevation of each of the convexities and concavities.
6. The method of claim 1 , wherein the feature vectors include the radius of curvature of each of the convexities and concavities.
7. The method of claim 1 , wherein the feature vectors features include the distance between adjacent convexities.
8. A system for controlling an autonomous vehicle, comprising:
an object classification module, including a processor, configured to:
receive lidar point cloud data associated with an object observed by a sensor system of an autonomous vehicle;
determine a bounding curve as an outline of the object in the lidar point cloud data and having a plurality of curve segments;
determine a feature vector for each of the plurality of curve segments based on convexities and concavities associated with the bounding curve; and
classify the object by applying the feature vectors to a machine learning model.
9. The system of claim 8 , wherein the machine learning model is an artificial neural network model.
10. The system of claim 8 , wherein the feature vectors include gradients between adjacent convexities and concavities.
11. The system of claim 8 , wherein the feature vectors include the elevation of each of the convexities and concavities.
12. The system of claim 8 , wherein the feature vectors include the radius of curvature of each of the convexities and concavities.
13. The system of claim 8 , wherein the feature vectors include the distance between adjacent convexities.
14. An autonomous vehicle, comprising:
at least one sensor that provides lidar point cloud data; and
a controller that, by a processor and based on the lidar point cloud data:
receives the lidar point cloud data associated with an object observed by the at least one sensor;
determines a bounding curve associated with the sensor data;
determines a feature vector for each of the plurality of curve segments based on convexities and concavities associated with the bounding curve; and
classifies the object by applying the feature vectors to a machine learning model.
15. The autonomous vehicle of claim 14 , wherein the machine learning model is an artificial neural network model.
16. The autonomous vehicle of claim 14 , wherein the feature vectors include gradients between adjacent convexities and concavities.
17. The autonomous vehicle of claim 14 , wherein the feature vectors include the elevation of each of the convexities and concavities.
18. The autonomous vehicle of claim 14 , wherein the feature vectors include the radius of curvature of each of the convexities and concavities.