Perception for point clouds

A computer system for detecting object motion in a point cloud captured by a sensor system of an ego vehicle, the computer system comprising:

• at least one memory configured to store computer-readable instructions;
• and
• at least one hardware processor coupled to the at least one memory and configured to execute the computer-readable instructions, which upon execution cause the at least one hardware processor to implement operations comprising:
• applying clustering to the point cloud, and thereby identifying at least one moving object cluster within the point cloud, the point cloud comprising time-stamped points captured over a non-zero accumulation window;
• determining a motion model for the moving object cluster, by fitting one or more parameters of the motion model to the time-stamped points of that cluster;
• wherein the point cloud is a radar point cloud, and each point has a measured Doppler component;
• wherein the parameters of the motion model comprise linear and angular velocity parameters defining predicted linear and angular velocity of an object in 2D or 3D space;
• and
• wherein, for each point of the moving object cluster, a predicted Doppler component is determined from the motion model based on the predicted linear and angular velocity of the object, the linear and angular velocity parameters fitted by comparing the predicted Doppler components with the measured Doppler components;
• and
• using the motion model to transform the time-stamped points of the moving object cluster to a common reference time to compensate for smearing effects caused by motion of the ego vehicle over the accumulation window.

A non-transitory computer readable medium embodying computer program instructions, the computer program instructions configured so as, when executed on one or more hardware processors, to implement operations comprising:

• applying clustering to a point cloud captured by a sensor system of an ego vehicle, and thereby identifying at least one moving object cluster within the point cloud, the point cloud comprising time-stamped points captured over a non-zero accumulation window;
• determining a motion model for the moving object cluster, by fitting one or more parameters of the motion model to the time-stamped points of that cluster;
• using the motion model to transform the time-stamped points of the moving object cluster to a common reference time to compensate for smearing effects caused by motion of the ego vehicle over the accumulation window;
• and
• applying a perception component to the transformed points of the moving object cluster to extract information about structure exhibited in the transformed points.