Computer-based tools and techniques for vehicle detection

1 . A vehicle analysis system comprising:

an automatic license plate recognition (ALPR) module programmed for receiving image data from at least one ALPR camera in association with a license plate of a target vehicle;

a context camera module programmed for receiving image data from at least one context camera that is distinct from the at least one ALPR camera in association with the target vehicle, the context camera image data depicting an operating environment of the target vehicle;

a geographical data module programmed for accessing map image data from at least one geographical data storage which stores a plurality of map definitions associated with a plurality of restricted zones; and

at least one artificial intelligence model and/or algorithm programmed for:

combining at least a portion of each of the received ALPR camera image data, the received context camera image data, and the geographical data storage map image data,

processing the received context camera image data to identify one or more scene elements associated with the restricted zones,

using at the least portion of the combined data including the identified scene elements for determining a location of the target vehicle in relation to at least one restricted zone, wherein the determining comprises determining whether the target vehicle is present within the at least one restricted zone, and

generating an enforcement determination indicative of a violation when the target vehicle is determined to be present within the at least one restricted zone.

2 . The system of claim 1 , further comprising:

a global positioning system (GPS) module programmed for processing position data associated with a position of a system vehicle; and

the artificial intelligence model and/or algorithm further programmed for combining at least a portion of the processed position data with the combined ALPR camera image data, the context camera image data, and the geographical data storage map image data.

3 . The system of claim 1 , further comprising:

an inertial measurement unit (IMU) module programmed for processing movement data and/or acceleration data associated with movement of a system vehicle; and

the artificial intelligence model and/or algorithm further programmed for combining at least a portion of the processed movement data with the combined ALPR camera image data, the context camera image data, and the geographical data storage map image data.

4 . The system of claim 1 , further comprising a vehicle localization module programmed for tracking at least one position of the target vehicle.

5 . The system of claim 1 , further comprising a vehicle communication protocol module programmed for connecting to a data bus of a system vehicle.

6 . The system of claim 5 , further comprising the vehicle communication protocol module programmed for:

receiving at least velocity data associated with the system data; and

the artificial intelligence model and/or algorithm further programmed for combining at least a portion of the received velocity data with the combined ALPR camera image data, the context camera image data, and the geographical data storage map image data.

7 . The system of claim 5 , further comprising the vehicle communication protocol module programmed for receiving J1939 protocol communications from the system vehicle.

8 . The system of claim 1 , further comprising a curb detection algorithm programmed for detecting a position of a street curb in response to at least video or image data obtained from the context camera.

9 . The system of claim 8 , further comprising the curb detection algorithm programmed for detecting the position of the street curb in response to:

comparing an expected curb position associated with projecting a curb line obtained from the map image data of the geographical data storage, and

an estimated system vehicle position.

10 . The system of claim 1 , further comprising a crosswalk detection algorithm programmed for detecting a crosswalk pattern in at least a portion of the received image data.

11 . The system of claim 10 , further comprising the crosswalk detection algorithm programmed for detecting the crosswalk pattern in response to comparing the crosswalk pattern detected in the received image data with a projected position of the crosswalk pattern associated with the map image data of the geographical data storage.

12 . The system of claim 1 , further comprising a camera-to-camera calibration algorithm programmed for calibrating the ALPR camera and the context camera with respect to each other.

13 . The system of claim 1 , further comprising a camera-to-GPS calibration algorithm programmed for provides a mapping between:

at least one global positioning system (GPS) map feature and GPS coordinates associated with the map feature, and

at least a portion of the image data received from the context camera.

14 . The system of claim 1 , further comprising:

the ALPR module programmed for detecting at least on vehicle license plate, and

an algorithm programmed for mapping plate coordinates of the detected license plate onto at least a portion of the image data received from the context camera.

15 . The system of claim 1 , further comprising a bounding module programmed to:

draw a bounding box around at least one detected object in the received image data, and

to generate a mask corresponding to a shape of the detected object.

16 . The system of claim 1 , further comprising a violation detection module programmed to automatically detect a violation, the violation detected module programmed for:

analyzing at least a portion of the data combined by the artificial intelligence model and/or algorithm, and

projecting image data associated with the restricted zone onto image data displaying the target vehicle.

17 . The system of claim 1 , further comprising a scene semantic segmentation model module programmed for detecting at least one street feature in the received image data.

18 . The system of claim 17 , further comprising a lane tracking algorithm programmed for:

tracking the street feature over a predetermined time period,

comparing the street feature to at least a portion of the map image data stored in the geographical data storage, and

determining which street feature is most accurately associated with the restricted zone.

19 . The system of claim 1 , further comprising the ALPR module programmed for:

detecting a pattern of numbers and/or letters provisionally identified as a license plate,

calculating an angle between different corners of the detected license plate, and

when the calculated angle is not within a predetermined valid range, discarding image data associated with the detected license plate.

20 . The system of claim 1 , further comprising a vehicle selection algorithm programmed for:

analyzing the image data received from the ALPR camera and the image data received from the context camera in association with the target vehicle license plate,

matching the target vehicle image data captured by the context camera to a corresponding license plate for the target vehicle shown in the image data captured by the ALPR camera.

21 . The system of claim 1 , further comprising a vehicle selection algorithm programmed for determining a bounding box that best matches the license plate in response to a vehicle type and/or a proximity of a projection of the license plate image data to the bounding box.

22 . The system of claim 1 , wherein the restricted zone comprises at least one of a bus lane, a bus stop, a parking area, and/or a bicycle lane.