MACHINE LEARNING BASED OBJECT DETECTION FOR TRANSIT SYSTEMS
Systems and methods for object detection through a fare gate in a transit system is disclosed. The object detection system includes a radar positioned at a first position and a reflective surface positioned at a second position of the fare gate, a machine learning (ML) engine, and a forensic engine. The radar emits a signal and generates a primary clustered point cloud from the signal reflected back from the object in a primary field-of-view (FOV). The reflective surface has a secondary FOV with a secondary clustered point cloud of the object. The ML engine extracts features of the object from the primary clustered point cloud and the secondary clustered point cloud and correlates the features with object profiles. The ML engine further determines that an object profile corresponds with an anomaly and generates a flag. The forensic engine captures media corresponding to the object associated with the flag.
1 . An object detection system to detect an object transiting through a fare gate in a transit system, the object detection system comprises:
a radar positioned at a first position of the fare gate, wherein the radar:
emits a signal, and
generates a primary point cloud from the signal reflected back from the object in a primary field-of-view (FOV);
clustering the primary point cloud to produce a primary clustered point cloud;
a reflective surface positioned at a second position of the fare gate, wherein:
the reflective surface has a secondary FOV, and
the secondary FOV has a secondary clustered point cloud of the object;
a machine learning (ML) engine, wherein the ML engine is operable to:
extract a plurality of features of the object from the primary clustered point cloud and the secondary clustered point cloud,
correlate the plurality of features with a plurality of object profiles, wherein a subset of the plurality of object profiles includes anomalies,
determine an object profile corresponding with an anomaly associated with the object, and
generate a flag upon identifying the anomaly; and
a forensic engine to capture media corresponding to the object associated with the flag.
2 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein the primary FOV and the secondary FOV are variable based on a position of the radar and a position of the reflective surface as a plurality of spherical coordinates.
3 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein the reflective surface is passive and comprises:
a plurality of copper layers and a substrate in between the plurality of copper layers; and
a plurality of wavelength-related spatially arranged unit-cell patches with a variable geometry in a phase pattern.
4 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein the reflective surface comprises a pattern that delivers a polarization insensitive response.
5 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein the reflective surface has a size that is a function of an operational frequency wavelength and transmit power of the radar and an aperture efficiency of the reflective surface that depends on the distance to the radar.
6 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein:
the radar emits the signal at a plurality of incidence angles,
the reflective surface reflects the signal at a plurality of reflection angles, and
the plurality of incidence angles and the plurality of reflection angles affect an electromagnetic response of the reflective surface.
7 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , further comprises a plurality of reflective surfaces and a plurality of radars.
8 . The object detection system to detect the object transiting through the fare gate in the transit system of claim 1 , wherein the ML engine is trained by:
sampling a plurality of primary clustered point clouds and a plurality of secondary clustered point clouds,
assessing the plurality of object profiles determined by the ML engine, and
analyzing errors and updating engine parameters based on feedback.
9 . An object detection method for detecting an object transiting through a fare gate in a transit system, the object detection method comprises:
positioning a radar at a first position of the fare gate, wherein the radar:
emits a signal, and
generates a primary clustered point cloud from the signal reflected back from the object in a primary field-of-view (FOV);
positioning a reflective surface at a second position of the fare gate, wherein:
the reflective surface has a secondary FOV, and the secondary FOV has a secondary clustered point cloud of the object;
configuring a machine learning (ML) engine to:
extract a plurality of features of the object from the primary clustered point cloud and the secondary clustered point cloud,
correlate the plurality of features with a plurality of object profiles, wherein a subset of the plurality of object profiles includes anomalies,
determine an object profile corresponding with an anomaly associated with the object, and
generate a flag upon identifying the anomaly; and
capturing media corresponding to the object associated with the flag via a forensic engine.
10 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein the primary FOV and the secondary FOV are variable based on a position of the radar and a position of the reflective surface as a plurality of spherical coordinates.
11 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein configuring the ML engine comprises training the ML engine by:
sampling a plurality of primary point clouds and a plurality of secondary point clouds,
assessing the plurality of object profiles determined by the ML engine, and
analyzing errors and updating engine parameters based on feedback.
12 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein:
the radar emits the signal at a plurality of incidence angles,
the reflective surface reflects the signal at a plurality of reflection angles, and
the plurality of incidence angles and the plurality of reflection angles affect an electromagnetic response of the reflective surface.
13 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein the reflective surface is passive and comprises:
a plurality of copper layers and a substrate in between the plurality of copper layers; and
a plurality of wavelength-related spatially arranged unit-cell patches with a variable geometry in a phase pattern.
14 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein the reflective surface comprises a pattern that delivers a polarization insensitive response.
15 . The object detection method for detecting the object transiting through the fare gate in the transit system of claim 9 , wherein the reflective surface has a size that is a function of an operational frequency wavelength of the radar and an aperture efficiency.
16 . A machine-readable medium having machine-executable instructions embodied thereon that, when executed by one or more processors, facilitate a method for object detection for detecting an object transiting through a fare gate in a transit system, wherein the method comprises:
positioning a radar at a first position of the fare gate, wherein the radar:
emits a signal, and
generates a primary point cloud from the signal reflected back from the object in a primary field-of-view (FOV);
positioning a reflective surface at a second position of the fare gate, wherein:
the reflective surface has a secondary FOV, and the secondary FOV has a secondary point cloud of the object;
configuring a machine learning (ML) engine to:
extract a plurality of features of the object from the primary point cloud and the secondary point cloud,
correlate the plurality of features with a plurality of object profiles, wherein a subset of the plurality of object profiles includes anomalies,
determine an object profile corresponding with an anomaly associated with the object, and
generate a flag upon identifying the anomaly; and
capturing media corresponding to the object associated with the flag via a forensic engine.
17 . The machine-readable medium, facilitating the method for object detection for detecting the object transiting through the fare gate in the transit system, of claim 16 , wherein the primary point cloud and/the secondary point cloud are processed with a clustering algorithm.
18 . The machine-readable medium, facilitating the method for object detection for detecting the object transiting through the fare gate in the transit system, of claim 16 , wherein configuring the ML engine comprises training the ML engine by:
sampling a plurality of primary point clouds and a plurality of secondary point clouds,
assessing the plurality of object profiles determined by the ML engine, and
analyzing errors and updating engine parameters based on feedback.
19 . The machine-readable medium, facilitating the method for object detection for detecting the object transiting through the fare gate in the transit system, of claim 16 , wherein:
the radar emits the signal at a plurality of incidence angles,
the reflective surface reflects the signal at a plurality of reflection angles, and
the plurality of incidence angles and the plurality of reflection angles affect an electromagnetic response of the reflective surface.
20 . The machine-readable medium, facilitating the method for object detection for detecting the object transiting through the fare gate in the transit system, of claim 16 , wherein the reflective surface is passive and comprises:
a plurality of copper layers and a substrate in between the plurality of copper layers; and
a plurality of wavelength-related spatially arranged unit-cell patches with a variable geometry in a phase pattern.