Autonomous vehicle system for performing object detections using a logistic cylinder pedestrian model

A system, comprising:

• an electronic device;
• and
• a computer-readable storage medium comprising one or more programming instructions that, when executed, cause the electronic device to:
• obtain, from one or more sensors of an autonomous vehicle, point cloud data that includes one or more data points, wherein each data point of the one or more data points has a point label associated therewith;
• define an optimized circle in a two-dimensional plane by:
• projecting the point cloud data onto the two-dimensional plane, wherein each data point in the point cloud data is represented by a two-dimensional location and the point label, and
• using the two-dimensional location of each data point to determine an optimized center parameter value of the optimized circle and an optimized radius parameter value of the optimized circle that, collectively, maximize a probability distribution of a center parameter and a radius parameter across the point cloud data;
• determine whether one or more of the data points of the point cloud data are located within the optimized circle;
• in response to determining that one or more of the data points of the point cloud data are located within the optimized circle, for each data point that is located within the optimized circle, assign a pedestrian class value to the point label of the data point;
• determine whether one or more of the data points of the point cloud data are located outside of the optimized circle;
• in response to determining that one or more of the data points of the point cloud data are located outside of the optimized circle, for each data point that is located outside of the optimized circle, assign a non-pedestrian class value to the point label of the data point;
• and
• use the point cloud data to control one or more driving operations of the autonomous vehicle.

A non-transitory computer-readable medium that stores instructions that, when executed by at least one computing device, will cause the at least one computing device to perform computing operations comprising:

• obtaining point cloud data that includes one or more data points, wherein each data point of the one or more data points has a point label associated therewith;
• defining an optimized circle in a two-dimensional plane by:
• projecting the data points of the point cloud data onto the two-dimensional plane, where each data point in the point cloud data is represented by a two-dimensional location and the point label;
• and
• using the two-dimensional location of each data point to determine an optimized center parameter value of the optimized circle and an optimized radius parameter value that, collectively, maximize a probability distribution of a center parameter and a radius parameter across the point cloud data;
• determining whether one or more of the data points of the point cloud data are located within the optimized circle;
• in response to determining that one or more of the data points of the point cloud data are located within the optimized circle, for each data point that is located within the optimized circle, assigning a pedestrian class value to the point label of the data point;
• determining whether one or more of the data points of the point cloud data are located outside of the optimized circle;
• in response to determining that one or more data points of the point cloud data are located outside of the optimized circle, for each data point that is located outside of the optimized circle, assigning a non-pedestrian class value to the point label of the data point;
• and
• using the point cloud data to control one or more driving operations of an autonomous vehicle.