Machine learned user interface for dynamic control of intelligent radar clutter mitigation
Systems and methods for sensing targets within an ambient environment. One example system includes a human machine interface and an electronic processor. The electronic processor is configured to receive sensor data including a plurality of targets. The electronic processor is configured to select a clutter mitigation model and generate, with the clutter mitigation model, a subset of the plurality of targets based on the sensor data. The electronic processor is configured to receive, from the human machine interface, an operator sensitivity setting. The electronic processor is configured to generate, with a machine learning mapping function, a post-processing parameter based on the operator sensitivity setting. The electronic processor is configured to process the subset of the plurality of targets based on the post-processing parameter to select, from the subset, a display target. The electronic processor is configured to display, via the human machine interface, the display target.
1 . A sensor system configured to sense targets within an ambient environment, the system comprising:
a human machine interface; and
an electronic processor coupled to the human machine interface and configured to:
receive sensor data including a plurality of targets;
select a clutter mitigation model;
generate, with the clutter mitigation model, a subset of the plurality of targets based on the sensor data;
receive, from the human machine interface, an operator sensitivity setting;
generate, with a machine learning mapping function, a post-processing parameter based on the operator sensitivity setting;
process the subset of the plurality of targets based on the post-processing parameter to select, from the subset of the plurality of targets, a display target; and
display, via the human machine interface, the display target;
wherein the post-processing parameter is a consistency counter threshold; and
processing the subset of the plurality of targets includes selecting, from the subset of the plurality of targets, the display target when the number of times that a confidence value for the display target exceeds a confidence threshold exceeds the consistency counter threshold.
2 . The system of claim 1 , wherein the electronic processor is further configured to:
select the clutter mitigation model and a second clutter mitigation model based on a model selection input received from the human machine interface;
generate, with the second clutter mitigation model, a second subset of the plurality of targets based on the sensor data;
process the second subset of the plurality of targets based on the post-processing parameter to select, from the subset of the plurality of targets, a second display target; and
display, via the human machine interface, the second display target.
3 . The system of claim 1 , wherein the electronic processor is further configured to:
retrieve, from a database, an operator profile for a current operator of the sensor system;
determine, based on the operator profile, an operator preference; and
generate the post-processing parameter based on the operator preference and the operator sensitivity setting.
4 . The system of claim 3 , wherein the operator preference is one selected from a group consisting of an operator preference for an ambient area type, an operator preference for the clutter mitigation model, an operator preference for a target type, and an operator preference for a sensor type for the sensor system.
5 . The system of claim 1 , wherein:
the post-processing parameter is a confidence threshold; and
the electronic processor is further configured to process the subset of the plurality of targets by selecting, from the subset of the plurality of targets, a display target when a confidence value for the display target exceeds the confidence threshold.
6 . The system of claim 1 , wherein the electronic processor is further configured to:
receive, via the human machine interface, operator feedback for the display target; and
train the machine learning mapping function using the operator feedback and the operator sensitivity setting.
7 . The system of claim 1 , wherein the electronic processor is further configured to select the clutter mitigation model based on the ambient environment.
8 . The system of claim 1 , wherein the clutter mitigation model is a machine learning model configured to identify clutter within the sensor data to identify which of the plurality of targets are true targets.
9 . The system of claim 1 , wherein the sensor system is a radar system.
10 . A method for operating a sensor system configured to sense targets within an ambient environment, the method comprising:
receiving sensor data including a plurality of targets;
selecting a clutter mitigation model;
generating, with the clutter mitigation model, a subset of the plurality of targets based on the sensor data;
receiving an operator sensitivity setting;
generating, with a machine learning mapping function, a post-processing parameter based on the operator sensitivity setting;
processing the subset of the plurality of targets based on the post-processing parameter to select, from the subset of the plurality of targets, a display target; and
displaying the display target;
wherein the post-processing parameter is a consistency counter threshold; and
wherein processing the subset of the plurality of targets includes selecting, from the subset of the plurality of targets, a display target when the number of times that a confidence value for the display target exceeds the confidence threshold exceeds the consistency counter threshold.
11 . The method of claim 10 , further comprising:
selecting the clutter mitigation model and a second clutter mitigation model based on a model selection input received from an operator of the system;
generating, with the second clutter mitigation model, a second subset of the plurality of targets based on the sensor data;
processing the second subset of the plurality of targets based on the post-processing parameter to select, from the subset of the plurality of targets, a second display target; and
displaying the second display target.
12 . The method of claim 10 , further comprising:
retrieving, from a database, an operator profile for a current operator of the sensor system;
determining, based on the operator profile, an operator preference; and
generating the post-processing parameter based on the operator preference and the operator sensitivity setting.
13 . The method of claim 12 , wherein determining the operator preference includes determining one selected from a group consisting of an operator preference for an ambient area type, an operator preference for the clutter mitigation model, an operator preference for a target type, and an operator preference for a sensor type for the sensor system.
14 . The method of claim 10 , wherein:
the post-processing parameter is a confidence threshold; and
processing the subset of the plurality of targets includes selecting, from the subset of the plurality of targets, a display target when a confidence value for the display target exceeds the confidence threshold.
15 . The method of claim 10 , further comprising:
receiving operator feedback for the display target; and
training the machine learning mapping function using the operator feedback and the operator sensitivity setting.
16 . The method of claim 10 , wherein selecting the clutter mitigation model includes selecting the clutter mitigation model based on the ambient environment.
17 . The method of claim 10 , wherein the clutter mitigation model is a machine learning model configured to identify clutter within the sensor data to identify which of the plurality of targets are true targets.
18 . The method of claim 10 , wherein the sensor system is a radar system.