Detecting phantom objects in ranging sensor data
Object detection for ranging sensor data may detect objects which are not actually present in the environment. To identify certain of these phantom objects, objects detected in the environment are analyzed to determine whether they are enclosed by another object and if the enclosed object has a distance from the ranging sensor higher than the enclosing object. This may suggest that the enclosing object has a surface or other feature that is sensed as additional depth that manifests as a separate detectable object. These phantom objects are identified and removed from further perception processing.
1 . A method for detecting phantom objects in ranging sensor data, comprising:
receiving ranging sensor data captured by a ranging sensor;
identifying a set of candidate objects in the ranging sensor data, each candidate object having respective positions in the ranging sensor data;
identifying an enclosed object in the set of candidate objects having a position at least partially enclosed by a position of an enclosing object in the set of candidate objects;
determining that the enclosed object has a further distance from the ranging sensor than the enclosing object;
removing or maintaining the enclosed object from the set of candidate objects in the ranging sensor data based as least in part in part on the determination that the enclosed object has the further distance than the enclosing object; and
determining relative movement of the enclosed object and the enclosing object with respect to one another across frames of the sensor data; and
wherein removing or maintaining of the enclosed object is further based on the relative movement.
2 . The method of claim 1 , wherein the enclosed object is removed or maintained further based on the type of the enclosing object.
3 . The method of claim 1 , wherein the position of the enclosed object is a bounding box that is enclosed by a bounding box of the enclosing object.
4 . The method of claim 1 , wherein the ranging sensor is a LIDAR sensor.
5 . The method of claim 1 , further comprising:
identifying another object in the set of candidate objects different than the enclosing or enclosed objects;
determining a relative movement of the enclosed object and the other object with respect to one another across frames of the sensor data; and
wherein removing or maintaining the enclosed object is further based on the relative movement.
6 . A system, comprising:
a processor; and
a non-transitory computer-readable storage medium containing instructions for execution by the processor for:
receiving ranging sensor data captured by a ranging sensor;
identifying a set of candidate objects in the ranging sensor data, each candidate object having respective positions in the ranging sensor data;
identifying an enclosed object in the set of candidate objects having a position at least partially enclosed by a position of an enclosing object in the set of candidate objects;
determining that the enclosed object has a further distance from the ranging sensor than the enclosing object;
removing or maintaining the enclosed object from the set of candidate objects in the ranging sensor data based as least in part in part on the determination that the enclosed object has the further distance than the enclosing object; and
determining relative movement of the enclosed object and the enclosing object with respect to one another across frames of the sensor data; and
wherein removing or maintaining of the enclosed object is further based on the relative movement.
7 . The system of claim 6 , wherein the enclosed object is removed or maintained further based on the type of the enclosing object.
8 . The system of claim 6 , wherein the position of the enclosed object is a bounding box that is enclosed by a bounding box of the enclosing object.
9 . The system of claim 6 , wherein the ranging sensor is a LIDAR sensor.
10 . The system of claim 6 , wherein the instructions are further executable by the processor for:
identifying another object in the set of candidate objects different than the enclosing or enclosed objects;
determining a relative movement of the enclosed object and the other object with respect to one another across frames of the sensor data; and
wherein removing or maintaining the enclosed object is further based on the relative movement.
11 . A non-transitory computer-readable medium containing instructions executable by one or more processors for:
receiving ranging sensor data captured by a ranging sensor;
identifying a set of candidate objects in the ranging sensor data, each candidate object having respective positions in the ranging sensor data;
identifying an enclosed object in the set of candidate objects having a position at least partially enclosed by a position of an enclosing object in the set of candidate objects;
determining that the enclosed object has a further distance from the ranging sensor than the enclosing object;
removing or maintaining the enclosed object from the set of candidate objects in the ranging sensor data based as least in part in part on the determination that the enclosed object has the further distance than the enclosing object; and
determining relative movement of the enclosed object and the enclosing object with respect to one another across frames of the sensor data; and
wherein removing or maintaining of the enclosed object is further based on the relative movement.
12 . The computer-readable medium of claim 11 , wherein the enclosed object is removed or maintained further based on the type of the enclosing object.
13 . The computer-readable medium of claim 11 , wherein the position of the enclosed object is a bounding box that is enclosed by a bounding box of the enclosing object.
14 . The computer-readable medium of claim 11 , wherein the ranging sensor is a LIDAR sensor.
15 . The computer-readable medium of claim 11 , wherein the instructions are further executable for:
identifying another object in the set of candidate objects different than the enclosing or enclosed objects;
determining a relative movement of the enclosed object and the other object with respect to one another across frames of the sensor data; and
wherein removing or maintaining the enclosed object is further based on the relative movement.