Method and system for identifying a vehicle of interest

Abstract

A method and system for identifying a vehicle of interest, the method comprising: accessing data corresponding to a vehicle of interest alert; accessing data from the sensor corresponding to the images of the one or more other vehicles; calculating, using a machine learning model, a vehicle-of-interest match estimate for the one or more other vehicles based at least in part on the data from the sensor; and calculating an identifier of the vehicle of interest.

Description

Technical Field

[0001] The present subject matter generally relates to systems and methods for identifying a vehicle of interest. Background Art

[0002] Alerts for vehicles of interest can be issued by law enforcement agencies to facilitate investigations. For example, an Amber Alert can be issued during a potential child abduction. An alert for a vehicle of interest can include a description of the vehicle of interest, which allows the public to report potential sightings of the vehicle of interest to the law enforcement agency.

[0003] Alerts for vehicles of interest often result in false sighting reports, where members of the public report sightings of vehicles that do not correspond to the vehicle of interest. Additionally, it can be difficult to communicate alerts for vehicles of interest to a large portion of the public. Thus, alerts for vehicles of interest can have low reliability. Further, while searching for a vehicle of interest, a driver may be distracted.

[0004] Therefore, improved systems and methods for identifying a vehicle of interest would be useful. Summary of the Invention

[0005] Generally, the present subject matter relates to systems and methods for detecting a vehicle of interest. For example, a computing device on a vehicle can access an alert for a vehicle of interest, which can correspond to a vehicle description provided by a law enforcement agency. The alert for the vehicle of interest can include one or more of a make, model, color, license plate number, number of passengers, age of passengers, and other identifying information of the vehicle of interest. One or more sensors on the vehicle can be configured to capture images of other vehicles operating in the vicinity of the vehicle. In an example arrangement, the sensor can include a camera (such as a front ADAS camera, a rear backup camera, a side view camera, etc.) that captures images of other vehicles. A machine learning model can analyze the images from the sensor to detect an estimate of a match of a vehicle of interest for other vehicles operating in the vicinity of the vehicle. For example, the machine learning model can be trained to detect a make, model, color, license plate number, number of passengers, age of passengers, and other features of other vehicles in the images. When a match with the vehicle of interest is identified, a vehicle of interest identification can be calculated. The vehicle of interest identification can include a make, model, color, license plate number, number of passengers, age of passengers, and other features of the other vehicle that matches the vehicle of interest. The vehicle of interest identification can also include an image of the other vehicle that matches the vehicle of interest and / or driving data of the other vehicle that matches the vehicle of interest, such as location, direction of travel, speed, etc. The vehicle of interest identification can be transmitted to a law enforcement agency to assist in the search for the vehicle of interest.

[0006] As can be seen above, the vehicle can advantageously assist in the search for a vehicle of interest, for example, while driving normally. Additionally, in an example arrangement, a machine learning model can automatically identify a vehicle of interest at least in part based on images of other vehicles captured by vehicle sensors. Thus, a vehicle of interest alert can be more reliable and safer than conventional approaches.

[0007] Aspects and advantages of the present disclosure will be set forth in part in the following description, or may be obvious from the description, or may be learned by practice.

[0008] In one example arrangement, a method for identifying a vehicle of interest includes: accessing, using a computing device on the vehicle, data corresponding to a vehicle of interest alert; accessing, using the computing device, data from sensors corresponding to images of one or more other vehicles; calculating, using a machine learning model on the computing device, a vehicle of interest match estimate for the one or more other vehicles at least in part based on the data from the sensors; calculating, in response to the vehicle of interest match estimate exceeding a threshold level, a vehicle of interest identification; and transmitting, using the computing device, data corresponding to the vehicle of interest identification to a remote computing device located external to the vehicle.

[0009] In another example arrangement, a system for identifying a vehicle of interest includes: a vehicle; sensors located on the vehicle; one or more processors located on the vehicle; and one or more non-transitory computer-readable media storing instructions that can be executed by the one or more processors to perform operations. The operations include: accessing data corresponding to a vehicle of interest alert; accessing data from the sensors corresponding to images of one or more other vehicles; calculating, using a machine learning model, a vehicle of interest match estimate for the one or more other vehicles at least in part based on the data from the sensors; calculating a vehicle of interest identification in response to the vehicle of interest match estimate exceeding a threshold level; and transmitting data corresponding to the vehicle of interest identification to a remote computing device located external to the vehicle.

[0010] These and other features, aspects, and advantages of the present disclosure will be better understood with reference to the following description and the appended claims. The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate arrangements of the present disclosure and, together with the description, serve to explain the principles of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The enabling disclosure of the present disclosure for a person of ordinary skill in the art is set forth in the specification with reference to the accompanying drawings.

[0012] Figure 1Is a side elevation view of a passenger vehicle arranged according to an example of the present subject matter.

[0013] Figure 2 Is Figure 1 A schematic diagram of the powertrain of an example vehicle.

[0014] Figure 3 Is arranged according to an example of the present subject matter Figure 1 A schematic diagram of an example control system of the vehicle.

[0015] Figure 4 Is a block diagram of certain components of a vehicle identification system of interest arranged according to an example of the present subject matter.

[0016] Figure 5 Is a schematic diagram of a vehicle identification system of interest arranged according to an example of the present subject matter.

[0017] Figure 6 Is a flowchart of a method for identifying a vehicle of interest arranged according to an example of the present subject matter. Detailed Description

[0018] Reference will now be made in detail to the arrangements of the present disclosure, one or more examples of which are shown in the accompanying drawings. Each example is provided by way of explanation of the present disclosure, and not by way of limitation of the present disclosure. Indeed, it will be apparent to those skilled in the art that various modifications and variations can be made in the present disclosure without departing from the scope or spirit thereof. For example, features shown or described as part of one arrangement can be used with another arrangement to yield yet another arrangement. Accordingly, the present disclosure is intended to cover such modifications and variations that fall within the scope of the appended claims and their equivalents.

[0019] As used herein, the term "comprising" is intended to be inclusive in a manner similar to the term "including". Similarly, the term "or" is generally intended to be inclusive (i.e., "A or B" is intended to mean "A or B or both"). As used throughout the specification and claims herein, approximating language is applied to modify any quantitative representation that could permissibly vary without resulting in a change in the basic function to which it is related. Accordingly, a value modified by one or several terms, such as "about", "approximately", and "substantially", is not limited to the precise value specified. In at least some instances, the approximating language can correspond to the precision of the instrument used to measure the value. For example, the approximating language can refer to within a ten percent (10%) margin.

[0020] Figure 1 Is a side elevation view of a passenger vehicle 100 arranged according to an example arrangement. Figure 2 Is a schematic diagram of the powertrain system 120 of the passenger vehicle 100. As Figure 1As shown, passenger vehicle 100 is shown as a sedan. However, Figure 1 the passenger vehicle 100 in Figure 1 is provided only as an example. For example, in an alternative exemplary arrangement, the passenger vehicle 100 can be a coupe, convertible, truck, van, sport utility vehicle, etc. Additionally, although described below in the context of the passenger vehicle 100, it will be understood that in other exemplary arrangements, the subject matter can be used in or with any other suitable vehicle, including commercial vehicles such as tractor-trailers, buses, box trucks, agricultural vehicles, construction vehicles, etc.

[0021] The passenger vehicle 100 can include a body 110 that rolls on wheels 116 during driving of the passenger vehicle 100. The body 110 defines an interior cabin 112, and a driver and passengers can enter the interior cabin 112 via doors 114 and sit on seats (not shown) within the interior cabin 112. Within the body 110, the passenger vehicle 100 can also include various systems for operating the passenger vehicle 100, including a motor system 122, a driveline system 124, an electrical energy storage system 126, etc.

[0022] Generally, the motor system 122, the driveline system 124, and the electrical energy storage system 126 can be constructed in any conventional manner. For example, the motor system 122 can include prime movers such as an electric machine system 140 and an internal combustion engine system 142 ( Figure 2 ), which are operable to propel the passenger vehicle 100. Thus, the passenger vehicle 100 can be referred to as a hybrid vehicle. The motor system 122 can be disposed within the body 110 and can be coupled to the driveline system 124. The driveline system 124 is disposed in the power flow between the motor system 122 of the passenger vehicle 100 and the wheels 116. In certain exemplary arrangements, a torque converter 128 can be disposed in the power flow between the internal combustion engine system 142 and the driveline system 124 within the driveline system 120. The driveline system 124 is operable to provide various speed and torque ratios between the input and output of the driveline system 124. Thus, for example, the driveline system 124 can provide a mechanical advantage to assist in propelling the passenger vehicle 100 via the motor system 122. A differential 129 can be disposed between the driveline system 124 and the wheels 116 to couple the driveline system 124 and the wheels 116 while also allowing relative rotation between the wheels 116 on opposite sides of the body 110.

[0023] The electric machine system 140 can be selectively capable of operating as a motor to propel the passenger vehicle 100 or as a generator to provide electrical power to, for example, the energy storage system 126 and other power-consuming devices of the passenger vehicle 100. Thus, for example, the electric machine system 140 can operate as a motor in certain operating modes of the passenger vehicle 100, and the electric machine system 140 can operate as a generator in other operating modes of the passenger vehicle 100. The electric machine system 140 can be arranged in the driveline system 120 in various configurations. For example, the electric machine system 140 can be provided as a module in the power flow path between the internal combustion engine system 142 and the transmission system 124. As another example, the electric machine system 140 can be integrated within the transmission system 124.

[0024] The energy storage system 126 can include one or more batteries, capacitors, etc. for storing electrical energy. The electric machine system 140 is coupled to the energy storage system 126 and is selectively operable to charge the energy storage system 126 when operating as a generator and to draw electrical power from the energy storage system 126 to propel the passenger vehicle 100 when operating as a motor.

[0025] A braking system (not shown) is operable to decelerate the passenger vehicle 100. For example, the braking system can include friction brakes configured to selectively reduce the rotational speed of the wheels 116. The braking system can also be configured as a regenerative braking system that converts the kinetic energy of the wheels 116 into electric current. The operation of the motor system 122, the transmission system 124, the energy storage system 126, and the braking system is known to those skilled in the art and is not described in detail herein for the sake of brevity.

[0026] Figure 3FIG. is a schematic diagram of certain components of a control system 130 adapted to be used with a passenger vehicle 100. Generally, the control system 130 is configured to control the operation of the passenger vehicle 100 and components therein. The control system 130 can facilitate the operation of the passenger vehicle 100 in various operating modes. For example, the control system 130 can be configured to operate the passenger vehicle 100 in any one of a conventional mode, an electric mode, a hybrid mode, and a regenerative mode. In the conventional mode, the passenger vehicle 100 is propelled only by the internal combustion engine system 142. In contrast, the passenger vehicle 100 is propelled only by the electric motor system 140 in the electric mode. Relative to the electric mode, the conventional mode can provide an extended operating range for the passenger vehicle 100, and the passenger vehicle 100 can be quickly refilled at a gas station to allow the passenger vehicle 100 to continue operating in the conventional mode. In contrast, relative to the conventional mode, emissions of the passenger vehicle 100 can be significantly reduced in the electric mode, and the fuel efficiency of the passenger vehicle 100 can be significantly increased in the electric mode compared to the conventional mode. In the hybrid mode, the passenger vehicle 100 can be propelled by both the electric motor system 140 and the internal combustion engine system 142. In the regenerative mode, the electric motor system 140 can charge, for example, the energy storage system 126, and the internal combustion engine system 142 can propel the passenger vehicle 100. The various operating modes of the passenger vehicle 100 are well known to those skilled in the art and are not described in detail herein for the sake of brevity.

[0027] As Figure 3 shown, the control system 130 includes one or more computing devices 132 having one or more processors 134 and one or more memory devices 136 (hereinafter referred to as "memory 136"). In certain example arrangements, the control system 130 can correspond to an electronic control unit (ECU) of the passenger vehicle 100. The one or more memories 136 store information accessible by the one or more processors 134, including instructions 138 that can be executed and data 139 that can be used by the one or more processors 134. The one or more memories 136 can be of any type capable of storing information accessible by the one or more processors 134, including computer-readable media. The memory is a non-transitory medium, such as a hard disk drive, a memory card, an optical disc, solid state, magnetic tape memory, etc. The one or more memories 136 can include different combinations of the foregoing, whereby different portions of the instructions and data are stored on different types of media. The one or more processors 134 can be any conventional processor, such as a commercially available CPU. Alternatively, the one or more processors 134 can be a specialized device, such as an ASIC or other hardware-based processor.

[0028] Instruction 138 can be any set of instructions to be executed directly (such as machine code) or indirectly (such as a script) by the one or more processors 134. For example, instruction 138 can be stored as computing device code on a computing device-readable medium of the one or more memories 136. In this regard, the terms "instruction" and "program" may be used interchangeably herein. Instruction 138 can be stored in an object code format for direct processing by a processor, or in any other computing device language (including a script or a collection of independent source code modules that are interpreted or pre-compiled as needed). Data 139 can be retrieved, stored, or modified by the one or more processors 134 according to instruction 138. For example, data 139 of the one or more memories 136 can store information from sensors of various systems of the passenger vehicle 100, the various systems including a motor system 122 (e.g., an electric motor system 140 and an internal combustion engine system 142), a driveline system 124, a power storage system 126, etc. In Figure 3 FIG., the processor 134, the memory 136, and other elements of the computing device 132 are shown within the same box. However, the computing device 132 can actually include multiple processors, computing devices, and / or memories that may or may not be stored within a common physical housing. Similarly, the one or more memories 136 can be a hard disk drive or other storage medium located in a housing different from the housing of the processor 134. Thus, the computing device 132 will be understood to include a processor and a collection of one or more memories that may or may not operate in parallel.

[0029] The computing device 132 may be configured to communicate with various components of the passenger vehicle 100. For example, the computing device 132 may be operatively communicable with various systems of the passenger vehicle 100, the various systems including a motor system 122 (e.g., an electric motor system 140 and an internal combustion engine system 142), a driveline system 124, an electrical energy storage system 126, and the like. For example, the computing device 132 may be operatively communicable with an engine control unit (ECU) (not shown) of the motor system 122 and a transmission control unit (TCU) (not shown) of the driveline system 124. The computing device 132 may also be operatively communicable with other systems of the passenger vehicle 100, the other systems including: a passenger / driver information system 150, e.g., the passenger / driver information system 150 includes one or more mode displays, speakers, gauges, etc. within the interior compartment 112 for providing information to the passenger / driver regarding the operation of the passenger vehicle 100; a cabin environment system 152 for modifying the temperature of the interior compartment 112, e.g., via air conditioning, heating, etc.; a navigation system 154 for navigating the passenger vehicle 100 to a destination; and / or a positioning system 156 for determining the current location of the passenger vehicle 100 (e.g., GPS coordinates). The computing device 132 may be configured to control systems 122, 124, 126 at least in part based on inputs received from an operator via a user interface (not shown), the user interface may include one or more of a steering wheel, an accelerator pedal, a clutch pedal, a brake pedal, a turn signal lever, a hazard warning light switch, etc.

[0030] The control system 130 may also include a wireless communication system 160 that facilitates wireless communication with other systems. For example, the wireless communication system 160 may wirelessly connect the control system 130 to one or more other vehicles, buildings, etc., directly or via a communication network. The wireless communication system 160 may include an antenna and a chipset configured to communicate according to one or more wireless communication protocols, such as the communication protocols described in Bluetooth, IEEE802.11, GSM, CDMA, UMTS, EV-DO, WiMAX, LTE, Zigbee, dedicated short range communication (DSRC), radio frequency identification (RFID) communication, etc. It should be understood that the internal communication between the computing device 132 and systems 122, 124, 126, 140, 142 within the passenger vehicle 100 may be wired and / or wireless. As an example, systems within the passenger vehicle 100 may be connected and communicate via a CAN bus.

[0031] As Figure 1As shown, the passenger vehicle 100 may include a front camera 158 and a rear camera 159. The front camera 158 may be a component of an advanced driver assistance system (ADAS) of the passenger vehicle 100. For example, the front camera 158 may be oriented on the vehicle body 110 along the forward direction of travel. Thus, the front camera 158 may capture an image of the area in front of the passenger vehicle 100 during travel. For example, the front camera 158 may capture an image of a vehicle in front of the passenger vehicle 100 during travel, and the image from the front camera 158 may be used for adaptive cruise control, forward collision warning, etc. The rear camera 159 may be a rearview camera for the passenger vehicle 100. For example, the rear camera 159 may be oriented on the vehicle body 110 along the reverse direction of travel. Thus, the rear camera 159 may capture an image of the area behind the passenger vehicle 100 during travel. For example, the rear camera 159 may capture an image of a vehicle or a pedestrian behind the passenger vehicle 100 during travel, and the image from the rear camera 159 may be presented on the display of the driver information system 150 such that the driver of the passenger vehicle 100 may utilize such an image during reverse travel of the passenger vehicle 100. The passenger vehicle 100 may also include other cameras, such as side cameras, to facilitate lane changes of the passenger vehicle 100.

[0032] Turning now to Figure 4 and 5 , the passenger vehicle 100 may also include features for detecting a vehicle of interest 310. Additionally, Figure 4 is a block diagram of a process of a vehicle of interest identification system 200 for a passenger vehicle 100 arranged according to an example of the present subject matter, and Figure 5 is a schematic diagram of a passenger vehicle 100 using the vehicle of interest identification system 200 to identify and locate a vehicle of interest 310. As an example, the process of the vehicle of interest identification system 200 may be implemented on the control system 130 via the processor 134 such that the process is performed at the edge or on the passenger vehicle 100. Thus, as Figure 5 shown, each passenger vehicle 100 may implement the vehicle of interest identification system 200 at the edge or on the passenger vehicle 100.

[0033] An interested vehicle identification system 200 of a passenger vehicle 100 can assist a law enforcement agency 330 or other emergency service provider in identifying and locating an interested vehicle 310. For example, the law enforcement agency 330 or other emergency service provider can issue an interested vehicle alert to facilitate an investigation, such as an Amber Alert or a criminal search. The interested vehicle alert can include a description of the interested vehicle 310, such as the make, model, color, license plate number, number of passengers, and ages of the passengers. The interested vehicle identification system 200 can receive the interested vehicle alert (and the description of the interested vehicle 310) via a remote computing device 202. In an example arrangement, a driver (or other operator or owner) of the passenger vehicle 100 can register or enroll the interested vehicle identification system 200 to receive interested vehicle alerts from the law enforcement agency 330 or other emergency service provider. As discussed in more detail below, the interested vehicle identification system 200 can be configured to identify and locate the interested vehicle 310 from among various other vehicles located around the passenger vehicle 100.

[0034] Reference Figure 4 , the interested vehicle identification system 200 of the passenger vehicle 100 can utilize images from sensors, such as one or more of a front camera 158, a rear camera 159, and / or a side view camera, to detect and identify the interested vehicle 310. Additionally, the interested vehicle identification system 200 can utilize images of other vehicles to identify and locate the interested vehicle 310 from among various other vehicles located around the passenger vehicle 100. When the interested vehicle 310 is identified, the interested vehicle identification system 200 can calculate an interested vehicle identifier to transmit to the law enforcement agency 330 or other emergency service provider to assist in the investigation of the interested vehicle 310.

[0035] In some example arrangements, the interested vehicle identification system 200 can be configured as part of a control system 130 of the passenger vehicle 100. Thus, for example, the computing device 132 can be programmed to implement the interested vehicle identification system 200 in some example arrangements. It will be understood that the interested vehicle identification system 200 can be used in or with any suitable vehicle. Thus, although described below in the context of the passenger vehicle 100, the interested vehicle identification system 200 can be used in or with an automobile, bus, truck, van, or any other vehicle traveling on a road. As described above, the interested vehicle identification system 200 can advantageously identify the interested vehicle at least in part based on images captured by sensors on the vehicle having the interested vehicle identification system 200.

[0036] Reference Figure 4, the vehicle-of-interest identification system 200 may include an image sensor 210 or communicate with the image sensor 210, and the image sensor 210 is configured to generate data corresponding to images of one or more other vehicles around the passenger vehicle 100. As an example, the image sensor 210 may include one or more of a side-view camera, a front camera 158, and a rear camera 159. Thus, for example, an image from the image sensor 210 may be a picture or video from the front camera 158, a picture or video from the rear camera 159, or a picture from a side-view camera. Accordingly, the vehicle-of-interest identification system 200 on the passenger vehicle 100 may access or receive data corresponding to images of other vehicles from the image sensor 210.

[0037] The vehicle-of-interest identification system 200 may further include a positioning system 220. The positioning system 220, which may correspond to the positioning system 156, may be configured to determine the current position (e.g., GPS coordinates) of the passenger vehicle 100. The positioning system 220 may be used to estimate the position of the vehicle of interest 310 when it is identified by the vehicle-of-interest identification system 200.

[0038] The vehicle-of-interest identification system 200 may further include a vehicle-of-interest identification model 230. The vehicle-of-interest identification model 230 may include a machine learning model that is configured to identify a vehicle of interest at least in part based on images from the image sensor 210. As an example, the vehicle-of-interest identification model 230 may utilize a convolutional neural network or other machine learning detection algorithms that are trained (first supervised and then unsupervised) to recognize, detect, or identify a vehicle of interest in images of other vehicles from the image sensor 210. For example, the vehicle-of-interest identification model 230 may calculate a vehicle-of-interest matching estimate for other vehicles based on data from the image sensor 210. The vehicle-of-interest matching estimate may correspond to the likelihood that another vehicle corresponds to the vehicle of interest 310 as calculated by the machine learning model.

[0039] The vehicle-of-interest identification model 230 may be trained using a training data set determined using information describing previous vehicles. The training data set may include one or more positive samples, where each positive sample represents a vehicle identification feature. The vehicle identification feature may include one or more of a brand, a model, a color, a license plate number, the number of passengers, and the age of the passengers. Thus, the machine learning model of the vehicle-of-interest identification model 230 may be trained to identify features of other vehicles in images from the image sensor 210.

[0040] The vehicle of interest identification model 230 can also calculate the vehicle of interest identifications of other vehicles around the passenger vehicle 100. In addition, for example, when the vehicle of interest identification model 230 identifies a vehicle of interest 310 from an image of the image sensor 210, the vehicle of interest identification model 230 can calculate the vehicle of interest identification. The vehicle of interest identification can include one or more of the position of the vehicle of interest 310, the driving direction of the vehicle of interest 310, the speed of the vehicle of interest 310, and the image of the vehicle of interest 310.

[0041] When the vehicle of interest match estimate for other vehicles exceeds a threshold, which can be selected depending on the desired matching likelihood, the vehicle of interest identification model 230 can generate a vehicle of interest identification. The vehicle of interest match estimate can correspond to the calculated likelihood that one of the other vehicles around the passenger vehicle 100 corresponds to the vehicle of interest 310.

[0042] The vehicle of interest identification system 200 can also include features for transmitting the vehicle of interest identification to other computing devices located outside the passenger vehicle 100. For example, the vehicle of interest identification system 200 can include a communication system 240 that is configured to transmit data with other computing devices (such as a remote computing device 202 that is separately located from the passenger vehicle 100). In an example arrangement, the communication system 240 can include or correspond to the wireless communication system 160. As an example, the remote computing device 202 can be a cloud server 320( Figure 5 ). For example, when the vehicle of interest identification model 230 identifies a potential match of the vehicle of interest 310 from an image of the image sensor 210, such as when the vehicle of interest match estimate for other vehicles exceeds a threshold, the vehicle of interest identification system 200 can be configured to transmit data corresponding to the vehicle of interest identification to the remote computing device 202.

[0043] The remote computing device 202 can be configured to alert a law enforcement agency 330 or other emergency service provider about a potential match of the vehicle of interest 310. In addition, the remote computing device 202 can be configured to transmit data corresponding to one or more of the position of the vehicle of interest 310, the driving direction of the vehicle of interest 310, the speed of the vehicle of interest 310, and the image of the vehicle of interest 310 to the law enforcement agency 330 or other emergency service provider. Thus, the vehicle of interest identification system 200 can assist the law enforcement agency 330 or other emergency service provider by transmitting location and other data related to the vehicle of interest 310 to the law enforcement agency 330 or other emergency service provider via the remote computing device 202. Then, the law enforcement agency 330 or other emergency service provider can investigate the vehicle of interest 310.

[0044] In an example arrangement, the remote computing device 202 can be a cloud server. Thus, for example, the communication system 240 can be configured for vehicle-to-infrastructure communication. In certain example arrangements, the remote computing device 202 can include one or more other passenger vehicles 100 having the vehicle of interest identification system 200. Thus, the communication system 240 can be configured for vehicle-to-vehicle communication. In such an arrangement, the vehicle of interest identification system 200 of the passenger vehicle 100 can be configured to coordinate the observation of the vehicle of interest 310. For example, one passenger vehicle 100 can cut off the observation of the vehicle of interest 310 based on the driving behavior of the vehicle of interest 310. Multiple passenger vehicles 100 can also jointly observe the vehicle of interest 310 to collect information about the vehicle of interest 310, for example, information that each passenger vehicle 100 cannot collect alone.

[0045] In an example arrangement, the vehicle of interest alert can include a geofence 340 for the vehicle of interest 310. For example, a law enforcement agency 330 or other emergency service provider can have the last known location and / or estimated location of the vehicle of interest 310. The geofence 340 can correspond to an area surrounding the last known location and / or estimated location of the vehicle of interest 310, within which the vehicle of interest identification system 200 in the passenger vehicle 100 is activated to assist in identifying and locating the vehicle of interest 310. Thus, for example, the passenger vehicle 100 outside the geofence 340 can be inactive and can not assist in identifying and locating the vehicle of interest 310. The size, shape, orientation, and other characteristics of the geofence 340 can be selected to facilitate the search for the vehicle of interest 310. For example, if the vehicle of interest 310 was last seen only a short time before the vehicle of interest alert, the geofence 340 can be small. Conversely, if the vehicle of interest 310 was last seen a long time before the vehicle of interest alert, the geofence 340 can be large. Similarly, if the vehicle of interest 310 is believed to be traveling quickly, the geofence 340 can be large. On the other hand, if the vehicle of interest 310 is believed to be traveling slowly, the geofence 340 can be small.

[0046] As Figure 5 shown, the geofence 340 can be shifted to an updated geofence 342. For example, one passenger vehicle 100 can identify and locate the vehicle of interest 310 within the original geofence 340 and can update the vehicle of interest alert with the updated geofence 342 to reflect the most recent information about the location of the vehicle of interest 310. It will be understood that the updating of the geofences 340, 342 can be repeated based on the most recent information about the vehicle of interest 310.

[0047] Figure 6FIG. 400 is a flow diagram of a method 400 for detecting a vehicle of interest arranged in accordance with an example of the present subject matter. Method 400 will generally be described with reference to a passenger vehicle 100 having a front camera 158 and a rear camera 159. For example, method 400 may be performed at least in part by a vehicle of interest identification system 200. However, method 400 may be suitable for use with any other suitable type of vehicle, control system configuration, and / or vehicle system. Additionally, although Figure 6 steps are depicted for purposes of illustration and discussion as being performed in a particular order, the methods and algorithms discussed herein are not limited to any particular order or arrangement. Those skilled in the art using the disclosure provided herein will understand that the various steps of the methods and algorithms disclosed herein can be omitted, rearranged, combined, and / or adapted in various ways without departing from the scope of the present disclosure.

[0048] At 410, data corresponding to a vehicle of interest alert may be accessed. For example, at 410, the vehicle of interest identification system 200 may access a vehicle of interest alert. A law enforcement agency 330 or other emergency service provider may generate a vehicle of interest alert along with a description of the vehicle of interest 310, such as make, model, color, license plate number, number of passengers, and ages of the passengers. At 410, the law enforcement agency 330 or other emergency service provider may transmit the vehicle of interest alert, and the passenger vehicle 100 may receive the vehicle of interest alert via the communication system 240. In an example arrangement, the vehicle of interest identification system 200 may access the vehicle of interest alert when the passenger vehicle 100 is registered or enrolled to receive vehicle of interest alerts from the law enforcement agency 330 or other emergency service provider. In an example arrangement, the vehicle of interest identification system 200 may access the vehicle of interest alert when the passenger vehicle 100 is located within a geofence 340. Conversely, when the passenger vehicle 100 is located outside of the geofence 340 and / or is not registered or enrolled to receive vehicle of interest alerts, the vehicle of interest identification system 200 may not access the vehicle of interest alert.

[0049] At 420, data corresponding to images of one or more vehicles may be accessed. For example, at 420, an image sensor may capture images of one or more vehicles in the vicinity of the vehicle. Additionally, at 420, the front camera 158, rear camera 159, and / or side cameras on the passenger vehicle 100 may capture images of one or more vehicles operating in the vicinity of the passenger vehicle 100. The control system 130 may access data corresponding to the images captured by the image sensor 210. For example, at 410, the vehicle of interest identification model 230 may access data corresponding to the images captured by the front camera 158, rear camera 159, and / or side cameras.

[0050] At 430, an interested vehicle matching estimate is calculated at least in part based on data corresponding to images of one or more vehicles from 420. For example, the interested vehicle identification model 230 may use, at 430, data corresponding to images captured by the front camera 158, the rear camera 159, and / or the side cameras at 420 to calculate the interested vehicle matching estimate. The interested vehicle identification model 230 may utilize a trained machine learning model to identify the interested vehicle 310 in the images captured by the image sensor 210, and the machine learning model may calculate the interested vehicle matching estimates for other vehicles in the images captured by the image sensor 210. The interested vehicle matching estimate may correspond to the likelihood that other vehicles around the passenger vehicle 100 match the interested vehicle 310 as calculated by the machine learning model.

[0051] Method 400 may further include calculating an interested vehicle identification, for example, where the interested vehicle identification corresponds to features of a potential match for the interested vehicle 310. For example, at 440, the interested vehicle identification model 230 may calculate the interested vehicle identification using one or more of the position of the interested vehicle 310, the direction of travel of the interested vehicle 310, the speed of the interested vehicle 310, and the image of the interested vehicle 310. In an example arrangement, when the interested vehicle identification model 230 identifies the interested vehicle 310 among other vehicles around the passenger vehicle 100 at 430, the interested vehicle identification model 230 may calculate the interested vehicle identification. As described above, the interested vehicle identification model 230 may utilize a trained machine learning model to identify the interested vehicle in images captured by the image sensor 210 (such as the front camera 158, the rear camera 159, and / or the side cameras).

[0052] At 440, data corresponding to the interested vehicle identification may be transmitted to a remote computing device. For example, the interested vehicle identification system 200 may transmit data corresponding to the interested vehicle identification to the remote computing device 202 via the communication system 240. For example, when the interested vehicle identification model 230 identifies a potential match for the interested vehicle 310 from the images of the image sensor 210, the interested vehicle identification system 200 may transmit, at 440, data corresponding to the interested vehicle identification.

[0053] Method 400 may further include alerting a law enforcement agency or other emergency service provider of a potential match for a vehicle of interest. For example, the remote computing device 202 may alert the law enforcement agency 330 or other emergency service provider of a potential match for the vehicle of interest 310. Additionally, the remote computing device 202 may transmit data corresponding to one or more locations of the vehicle of interest 310, the direction of travel of the vehicle of interest 310, the speed of the vehicle of interest 310, and an image of the vehicle of interest 310 to the law enforcement agency 330 or other emergency service provider. The law enforcement agency 330 or other emergency service provider may then investigate the vehicle of interest 310.

[0054] As can be seen from above, the present subject matter may provide a system and method for identifying a vehicle of interest that can assist an emergency service provider in locating the vehicle of interest. The vehicle of interest may be automatically identified by a machine learning model that analyzes images from one or more on-vehicle cameras. Participating vehicles may register to receive vehicle-of-interest alerts, and the vehicles may utilize a system at the edge to identify the vehicle of interest among various other vehicles surrounding the participating vehicle. When detected, the vehicle may report the vehicle of interest to the emergency service provider. Participation of the vehicle may also be determined by location (e.g., proximity to the vehicle of interest).

[0055] This written description uses examples to disclose the present subject matter and also enables any person skilled in the art to practice the present disclosure, including making and using any device or system and performing any incorporated method. The patentable scope of the present disclosure is defined by the claims and may include other examples that occur to those skilled in the art. If such other examples include structural elements that are different from the literal language of the claims, or if they include equivalent structural elements that are not materially different from the literal language of the claims, then they are intended to be within the scope of the claims. Example arrangements

[0056] First example arrangement: A method for identifying a vehicle of interest, comprising: accessing, using a computing device on a vehicle, data corresponding to a vehicle-of-interest alert; accessing, using the computing device, data from sensors corresponding to images of one or more other vehicles; calculating, using a machine learning model on the computing device, an estimate of a vehicle-of-interest match for the one or more other vehicles, at least in part based on the data from the sensors; calculating, in response to the vehicle-of-interest match estimate exceeding a threshold level, a vehicle-of-interest identification using the computing device; and transmitting, using the computing device, data corresponding to the vehicle-of-interest identification to a remote computing device located external to the vehicle.

[0057] Second example arrangement: The method according to the first example arrangement, wherein the vehicle-of-interest alert is issued by an emergency service agency.

[0058] Third exemplary arrangement: A method according to the first exemplary arrangement or the second exemplary arrangement, wherein the vehicle of interest alert includes data corresponding to one or more of the make, model, color, and license plate number of the vehicle of interest.

[0059] Fourth exemplary arrangement: The method according to any one of the first exemplary arrangement to the third exemplary arrangement, wherein the sensor includes one or more of an advanced driver assistance system camera, a rearview camera, and a side view camera.

[0060] Fifth exemplary arrangement: The method according to any one of the first exemplary arrangement to the fourth exemplary arrangement, wherein the matching of the vehicle of interest corresponds to the likelihood of one or more other vehicles calculated by the machine learning model matching the vehicle of interest.

[0061] Sixth exemplary arrangement: The method according to any one of the first exemplary arrangement to the fifth exemplary arrangement, wherein the machine learning model has been trained using a training data set determined using information describing previous vehicles, the training data set including one or more positive samples, each positive sample representing a vehicle identification feature.

[0062] Seventh exemplary arrangement: The method according to any one of the first exemplary arrangement to the sixth exemplary arrangement, wherein the vehicle identification feature includes one or more of make, model, color, and license plate number.

[0063] Eighth exemplary arrangement: The method according to any one of the first exemplary arrangement to the seventh exemplary arrangement, wherein the vehicle of interest identification includes one or more of the following: the location of the vehicle of interest; the direction of travel of the vehicle of interest; the speed of the vehicle of interest; and an image of the vehicle of interest.

[0064] Ninth exemplary arrangement: The method according to any one of the first exemplary arrangement to the eighth exemplary arrangement, wherein the remote computing device includes a cloud server, and the method further includes transmitting the vehicle of interest identification to an emergency service agency using the cloud server.

[0065] Tenth exemplary arrangement: The method according to any one of the first exemplary arrangement to the ninth exemplary arrangement, further including calculating, using the computing device, whether the vehicle is within a geofence for the vehicle of interest alert, wherein accessing the data from the sensor includes accessing the data from the sensor when the vehicle is within the geofence.

[0066] Eleventh example arrangement: The method according to any one of the first example arrangement to the tenth example arrangement further includes: using the computing device to access data corresponding to an updated vehicle alert of interest; using the computing device to calculate whether the vehicle is within an updated geofence for the updated vehicle alert of interest; and when the vehicle is within the updated geofence, continue to access the data from the sensor.

[0067] Twelfth example arrangement: The method according to any one of the first example arrangement to the eleventh example arrangement further includes: when the vehicle is outside the geofence, terminate the calculation of the vehicle of interest matching estimate.

[0068] Thirteenth example arrangement: A system for identifying a vehicle of interest includes: a vehicle; a sensor located on the vehicle; one or more processors located on the vehicle; and one or more non-transitory computer-readable media storing instructions that can be executed by the one or more processors to perform operations. The operations include: accessing data corresponding to a vehicle alert of interest; accessing data from the sensor corresponding to images of one or more other vehicles; at least partially based on the data from the sensor, using a machine learning model to calculate a vehicle of interest matching estimate for the one or more other vehicles; in response to the vehicle of interest matching estimate exceeding a threshold level, calculating a vehicle of interest identifier; and transmitting data corresponding to the vehicle of interest identifier to a remote computing device located outside the vehicle.

[0069] Fourteenth example arrangement: The system according to the thirteenth example arrangement, wherein the vehicle alert of interest includes data corresponding to one or more of the make, model, color, and license plate number of the vehicle of interest.

[0070] Fifteenth example arrangement: The system according to any one of the thirteenth or fourteenth example arrangements, wherein the sensor includes one or more of an advanced driver assistance system camera, a rearview camera, and a side view camera.

[0071] Sixteenth example arrangement: The system according to any one of the thirteenth example arrangement to the fifteenth example arrangement, wherein the vehicle of interest matching corresponds to the likelihood of the one or more other vehicles matching the vehicle of interest calculated by the machine learning model.

[0072] Seventeenth example arrangement: The system is arranged according to any one of the arrangements of the thirteenth to sixteenth example arrangements, wherein the vehicle identification of interest includes one or more of the following: the position of the vehicle of interest; the driving direction of the vehicle of interest; the speed of the vehicle of interest; and an image of the vehicle of interest.

[0073] Eighteenth example arrangement: The system is arranged according to any one of the arrangements of the thirteenth example arrangement to the seventeenth example arrangement, wherein the operation includes calculating whether the vehicle is within a geofence for the vehicle alert of interest, and wherein accessing the data from the sensors includes accessing the data from the sensors when the vehicle is within the geofence.

[0074] Nineteenth example arrangement: The system is arranged according to any one of the arrangements of the thirteenth to eighteenth example arrangements, wherein the operation includes: accessing data corresponding to an updated vehicle alert of interest; calculating whether the vehicle is within an updated geofence for the updated vehicle alert of interest; and continuing to access the data from the sensors when the vehicle is within the updated geofence.

[0075] Twentieth example arrangement: The system is arranged according to any one of the arrangements of the thirteenth to nineteenth example arrangements, wherein the operation includes: terminating the calculation of the vehicle match estimate of interest when the vehicle is outside the geofence.

Claims

1. A method for identifying a vehicle of interest, comprising: Accessing, using a computing device on the vehicle, data corresponding to a vehicle alert of interest; accessing, with the computing device, data from a sensor corresponding to images of one or more other vehicles; computing, using a machine learning model on the computing device, a vehicle of interest match estimate for the one or more other vehicles based at least in part on the data from the sensor; calculating, with the computing device, a vehicle of interest identification in response to the vehicle of interest match estimate exceeding a threshold level; as well as Data corresponding to the vehicle identification of interest is transmitted, using the computing device, to a remote computing device located external to the vehicle.

2. The method according to claim 1, wherein: The vehicle of interest alert is issued by emergency services.

3. The method according to claim 1, wherein: The vehicle of interest alert includes data corresponding to one or more of a make, model, color, and license plate number of the vehicle of interest. 4 . The method of claim 1 , wherein the sensor comprises one or more of an advanced driver assistance system camera, a backup camera, and a side view camera.

5. The method according to claim 1, wherein: The vehicle-of-interest match corresponds to a likelihood, calculated by the machine learning model, that the one or more other vehicles match the vehicle of interest.

6. The method according to claim 1, wherein: The machine learning model has been trained using a training data set, the training data set being determined using information describing previous vehicles, the training data set comprising one or more positive samples, each positive sample representing a vehicle identification feature.

7. The method of claim 6, wherein the vehicle identification feature comprises one or more of a make, a model, a color, and a license plate number.

8. The method of claim 1, wherein the vehicle identification of interest comprises one or more of the following: the location of the vehicle of interest; the direction of travel of the vehicle of interest; The speed of the vehicle of interest; and An image of the vehicle of interest.

9. The method of claim 1, wherein the remote computing device comprises a cloud server, The method further comprises: The vehicle identification of interest is transmitted to emergency services using the cloud server.

10. The method according to claim 1, further comprising: calculating, using the computing device, whether the vehicle is within a geofence for the vehicle of interest alert, Wherein, accessing the data from the sensor includes: accessing the data from the sensor when the vehicle is within the geo-fence.

11. The method according to claim 10, further comprising: accessing, using the computing device, data corresponding to an updated vehicle alert of interest; calculating, using the computing device, whether the vehicle is within an updated geo-fence for the updated vehicle of interest alert; as well as While the vehicle is within the updated geo-fence, the data from the sensor continues to be accessed.

12. The method according to claim 10, further comprising: When the vehicle is outside the geo-fence, calculation of the vehicle of interest match estimate is terminated.

13. A system for identifying a vehicle of interest, comprising: vehicle; a sensor located on said vehicle; one or more processors located on the vehicle; as well as One or more non-transitory computer-readable media storing instructions executable by the one or more processors to perform operations comprising: accessing data corresponding to vehicle alerts of interest; accessing data from the sensor corresponding to images of one or more other vehicles; calculating, using a machine learning model, a vehicle of interest match estimate for the one or more other vehicles based at least in part on the data from the sensor; In response to the vehicle of interest match estimate exceeding a threshold level, calculating a vehicle of interest identification; and Data corresponding to the vehicle identification of interest is transmitted to a remote computing device located external to the vehicle.

14. The system according to claim 13, wherein: The vehicle of interest alert includes data corresponding to one or more of a make, model, color, and license plate number of the vehicle of interest.

15. The system of claim 13, wherein the sensor comprises one or more of an advanced driver assistance system camera, a backup camera, and a side view camera.

16. The system of claim 13, wherein: The vehicle-of-interest match corresponds to a likelihood, calculated by the machine learning model, that the one or more other vehicles match the vehicle of interest.

17. The system of claim 13, wherein the vehicle identification of interest comprises one or more of the following: the location of the vehicle of interest; the direction of travel of the vehicle of interest; The speed of the vehicle of interest; and An image of the vehicle of interest.

18. The system of claim 13, wherein the operations comprise: calculating whether the vehicle is within a geo-fence for the vehicle of interest alert, Wherein accessing the data from the sensor comprises: accessing the data from the sensor when the vehicle is within the geo-fence.

19. The system of claim 18, wherein: The operations include: accessing data corresponding to updated vehicle alerts of interest; calculating whether the vehicle is within an updated geo-fence for the updated vehicle of interest alert; and While the vehicle is within the updated geo-fence, the data from the sensor continues to be accessed.

20. The system of claim 18, wherein the operations comprise: When the vehicle is outside the geo-fence, calculation of the vehicle of interest match estimate is terminated.

Images