Control system of AI (artificial intelligence) locomotive

Abstract

The invention provides a control system of an AI (artificial intelligence) locomotive, which comprises a locomotive starting battery, an AI device, a first license plate image acquisition device, a second license plate image acquisition device, a display device, a heat dissipation device and a power supply management device. Therefore, license plate image acquisition control, license plate detection, license plate recognition operation control and power supply management control are realized. Therefore, when the AI locomotive is ridden for patrol, the effects of license plate recognition and active detection are achieved, and the effects of good investigation maneuverability and investigation instantaneity are achieved.

Description

Technical Field

[0001] The present invention relates to a control system of an AI artificial intelligence motorcycle, and more particularly to a control system of an AI artificial intelligence motorcycle that can be mounted on an existing motorcycle and has the effects of license plate recognition and active detection when the motorcycle is patrolled, thereby achieving better detection and prevention mobility and instant detection and prevention effects. Background Art

[0002] Generally, police officers usually ride motorcycles when on patrol duty. When they find violators or suspected violators, they will stop them for questioning. Since this method of stopping and checking is time-consuming, if the police introduce technology law enforcement methods when they are on patrol duty, it will effectively avoid the above-mentioned shortcomings and have the functions of mobile detection and prevention and real-time detection and prevention in actual implementation.

[0003] The present invention aims to address the deficiencies of many existing police motorcycles and develops a multifunctional intelligent detection and prevention mobile platform. AI computers, cameras and other equipment are installed on the mobile platform, which can instantly recognize license plates of vehicles moving relatively to the moving vehicles or vehicles parked on the roadside, stolen vehicles or vehicles involved in crimes. Face recognition system software is installed on the AI ​​computer, and the police officers on duty wear portable and lightweight wireless camera devices to transmit images to the AI ​​computer on the motorcycle within the signal receiving range to perform instant face recognition function, immediately identify and warn wanted criminals or dangerous persons, and play an active role in investigation and protection of law enforcement personnel. In addition, an algorithm is developed using the portable camera to analyze image recognition and identify physiological changes caused by drinking as the object of alcohol detection. Radar speed measuring equipment can also be installed on motorcycles, combined with the on-board AI computer license plate recognition function to issue speeding vehicle warnings at certain points; and combined with cloud vehicle database information, speeding objects can be given instant mobile phone text message reminders to minimize casualties caused by speeding and achieve the goal of a friendly city. Summary of the invention

[0004] The main purpose of the present invention is to avoid the above-mentioned conventional deficiencies from continuing to exist, so that cameras and AI computers can be installed on existing motorcycles to achieve license plate recognition and active detection when patrolling on motorcycles, thereby achieving better detection and prevention mobility and real-time detection and prevention effects.

[0005] To achieve the above-mentioned purpose, the present invention is a control system for an AI locomotive, which includes: a locomotive provided with a locomotive starting battery, an AI device electrically connected to the locomotive starting battery, a first license plate image capture device electrically connected to the AI ​​device, a second license plate image capture device electrically connected to the AI ​​device, a display device electrically connected to the AI ​​device, a heat dissipation device electrically connected to the AI ​​device, and a power supply management device electrically connected to the AI ​​device. The AI ​​device and the heat dissipation device are arranged in the accommodation compartment of the locomotive, the first license plate image capture device and the second license plate image capture device are installed on the front of the locomotive, and the display device is arranged above the dashboard of the front of the locomotive; the power supply management device at least executes the opening step of the AI ​​device, the closing step of the AI ​​device and the charging step of the locomotive starting battery; when executing the opening step of the AI ​​device, the locomotive power switch provided for the locomotive starting battery is turned on, and the reset switch of the power supply management device is activated to turn on the locomotive starting battery to output power to the AI ​​device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device, and the display device displays the boot procedure and the system screen to complete the opening of the AI ​​device; when executing the closing step of the AI ​​device, the locomotive power switch is turned off, and the AI ​​device enters the shutdown procedure, and waits for the display device to display the shutdown After the startup is completed, the reset switch is activated again. When the display device has no output image, the AI ​​artificial intelligence device is closed. When the charging step of the locomotive starting battery is executed, the charging selection switch of the power supply management device is activated to the vehicle charging side position, so that the power supply management device receives the electric energy of the locomotive generator module of the locomotive to provide power to the input end of the power supply management device for charging the locomotive starting battery, and indirectly or directly supplies the power required for the operation of the AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device; in addition, the charging selection switch of the power supply management device can also be activated to the external charging side position, so that the Bluetooth communication charger of the power supply management device receives the alternating current from the external power supply as the charging of the locomotive starting battery, so as to indirectly or directly supply the power required for the operation of the AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device.

[0006] The present invention provides another control system of an AI artificial intelligence motorcycle, which includes: a motorcycle; an AI artificial intelligence device, which is arranged in the accommodation compartment of the motorcycle; a first license plate image capture device, which is electrically connected to the AI ​​artificial intelligence device, the first license plate image capture device includes a first image capture card device, the first image capture card device is installed on the front of the motorcycle, the first image capture card device is electrically connected to the first license plate image capture device through a coaxial cable line transmission signal through an analog image signal input end, so that the digital image signal output end of the first license plate image capture device is connected to the AI ​​through a wire transmission signal. The AI ​​device is electrically connected; a second license plate image capture device, which is electrically connected to the AI ​​device, the second license plate image capture device includes a second image capture card device, the second image capture card device is installed on the front of the motorcycle, the second image capture card device is electrically connected to the second license plate image capture device through a coaxial cable to transmit a signal through an analog image signal input end, so that the digital image signal output end of the second license plate image capture device is electrically connected to the AI ​​device through a wire transmission signal; a display device, which is electrically connected to the AI ​​device, the display device includes an image A / D PCB board / touch control board, the display device is arranged above the dashboard of the front of the vehicle, the image A / D PCB board / touch control board is arranged in the accommodating compartment, and the image A / D PCB board / touch control board is electrically connected to the display device through a bonding line; a wireless network AP device (router), which is electrically connected to the AI ​​device through an RJ45 port network cable or a USB port, or directly by a 5G built-in AI device casing A Wifi network module is wired to the antenna base of the computer case and then led to the outside of the vehicle body as a point-to-point wireless connection between the AI ​​artificial intelligence device and external devices; a radar speed measurement module is electrically connected to the AI ​​artificial intelligence device and is installed in the rear box of the motorcycle. The radar speed measurement module is connected to the RS485 port of the AI ​​artificial intelligence device through a UPT signal line, or can be connected to the RS232 port of the AI ​​artificial intelligence device through an RS485 / RS232 converter, or can be connected to the USB port of the AI ​​artificial intelligence device through an RS485 / USB converter. When the motorcycle is parked on the roadside, the radar speed measurement module cooperates with the license plate recognition function of the license plate image capture devices, and can instantly calculate the speed of each vehicle in motion and the matching license plate number through an image analysis algorithm;An RFID active receiver, which is electrically connected to the AI ​​device and is disposed on the front of the motorcycle. The RFID active receiver is connected to the RS485 port of the AI ​​device via a UPT signal line, or can be connected to the RS232 port of the AI ​​device via an RS485 / RS23 converter, or can be connected to the USB port of the AI ​​device via an RS485 / USB converter. The RFID active receiver can receive ETC / eTag chips on parking vehicles charged for roadside parking while the motorcycle is moving, and cooperate with the license plate recognition function of the license plate image capture devices. Through the image analysis algorithm, the speed of each vehicle in motion and the matching license plate number can be calculated in real time. ;

[0007] The present invention provides another license plate recognition control system for an AI artificial intelligence motorcycle, which includes: a motorcycle provided with a motorcycle starting battery, an AI artificial intelligence device electrically connected to the motorcycle starting battery, a first license plate image capture device electrically connected to the AI ​​artificial intelligence device, and a second license plate image capture device electrically connected to the AI ​​artificial intelligence device, the AI ​​artificial intelligence device is arranged in a storage compartment of the motorcycle, and the first license plate image capture device and the second license plate image capture device are installed on the front of the motorcycle; the AI ​​artificial intelligence device cooperates with the first license plate image capture device and the second license plate image capture device to perform license plate recognition operations; and the license plate recognition operations at least include the following steps:

[0008] Step 1: License plate image input, first obtain at least one license plate image from the first license plate image capture device and the second license plate image capture device and input it into the AI ​​artificial intelligence device.

[0009] Step 2: License plate detection, based on a neural network-like license plate recognition algorithm, the license plate frame position, size and tilt angle of the characters on the license plate in the license plate image are detected, and the detected license plate is cut out from the license plate image.

[0010] Step 3: Rotation correction: according to the tilt classification angle of the characters on the license plate determined in step 2, the license plate in the license plate image is rotated and corrected so that the characters on the license plate do not overlap as much as possible in the horizontal straight line position.

[0011] Step 4: License plate number recognition: input the rotation-corrected license plate image into the license plate recognition neural network to recognize the license plate number.

[0012] In the control system of the above-mentioned AI artificial intelligence locomotive, the locomotive power switch is further electrically connected to at least one protection device, a locomotive load, and a locomotive voltage and current stabilizing device, and the locomotive generator module is electrically connected to the locomotive voltage and current stabilizing device.

[0013] In the control system of the above-mentioned AI motorcycle, the AI ​​device is further electrically connected to a radar speed measurement module and a wireless network AP device (router) or the AI ​​device has a built-in 5G Wifi module.

[0014] In the control system of the above-mentioned AI locomotive, the wireless network AP device (router) or the AI ​​device 20 has a built-in 5G Wifi module as a wireless connection between the AI ​​device and other external devices.

[0015] In the control system of the above-mentioned AI artificial intelligence motorcycle, the AI ​​artificial intelligence device is electrically connected to a temperature and voltage detection device. The temperature and voltage detection device can communicate with the AI ​​artificial intelligence device in a two-way manner through a USB port and interface software. The temperature and voltage detection device detects the temperature / voltage / current values ​​of the first license plate image capture device, the second license plate image capture device, and the display device, and compares them with the safety values ​​set by the AI ​​artificial intelligence device, thereby driving the heat dissipation device to inlet and exhaust air to form a strong convection in the computer compartment to reduce the temperature.

[0016] In the control system of the above-mentioned AI locomotive, the AI ​​device is electrically connected to a speaker / buzzer, which is arranged on the display device. The triggering / signal amplification function of the speaker / buzzer is activated by connecting the amplifier component embedded in the image A / D PCB board / touch control board through the signal control cable, and the speaker / buzzer is connected to the GPIO / RS232 adapter of the AI ​​device through the UTP signal line, or the buzzer is connected to the speaker hole on the AI ​​device, and the timing of the sending mechanism is controlled by the AI ​​device.

[0017] In the control system of the above-mentioned AI artificial intelligence motorcycle, the digital image signal output end of the first license plate image capture device is electrically connected to the AI ​​artificial intelligence device via a USB3.0 cable transmission signal, and the digital image signal output end of the second license plate image capture device is electrically connected to the AI ​​artificial intelligence device via a USB3.0 cable transmission signal.

[0018] In the control system of the above-mentioned AI artificial intelligence motorcycle, the AI ​​artificial intelligence device is electrically connected to a three-button control switch, which is arranged on the handle of the motorcycle and is electrically connected to the AI ​​artificial intelligence device through a control line. The three-button control switch is a single-loop switching switch, and the control panel of the three-button control switch is arranged in the AI ​​artificial intelligence device with a USB port. The first button switch of the three-button control switch controls the temporary power supply or power supply cutoff of the display device through the AI ​​artificial intelligence device; the second button switch of the three-button control switch starts a slow shutdown program action through the AI ​​artificial intelligence device to shut down; the third button switch of the three-button control switch controls whether the image stream code of the first license plate image capture device enters the recognition program and whether the image stream code of the second license plate image capture device enters the recognition program through the built-in software of the AI ​​artificial intelligence device.

[0019] In the above-mentioned license plate recognition control system of another AI artificial intelligence motorcycle, the depthwise separable convolution of the neural network-like license plate recognition algorithm in step 2 gradually extracts the features of the license plate in the license plate image from 512 by 512 to a smaller size, and performs prediction through the prediction layers in six depths.

[0020] In the above-mentioned license plate recognition control system of another AI artificial intelligence motorcycle, the size of the first prediction layer is 32 times 32, the second prediction layer is 16 times 16, the third prediction layer is 8 times 8, the fourth prediction layer is 4 times 4, the fifth prediction layer is 2 times 2, and the sixth prediction layer is 1 times 1. After obtaining the license plate position and tilt angle classification in the license plate image, it is cut from the scene according to the coordinates and rotation correction is performed to complete the preprocessing of license plate recognition.

[0021] In the above-mentioned license plate recognition control system of another AI artificial intelligence motorcycle, the rotation angle in step 2 is divided into 8 categories, namely two groups of left and right angles of 0 degrees to 10 degrees, two groups of left and right angles of 10 degrees to 20 degrees, two groups of left and right angles of 20 degrees to 30 degrees, and two groups of left and right angles above 30 degrees, so as to cut the detected license plate out of the license plate image. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematic diagram of a control system block diagram of a first embodiment of the present invention.

[0023] Figure 2 Schematic diagram of a control system block diagram of a second embodiment of the present invention.

[0024] Figure 3 It is a three-dimensional appearance schematic diagram of the AI ​​artificial intelligence locomotive of the present invention.

[0025] Figure 4 It is a flowchart of a license plate recognition control system according to the third embodiment of the present invention.

[0026] Figure 5 It is a flow chart of the license plate recognition algorithm of the present invention.

[0027] Figure 6 It is a classification diagram of the inclination angle of the license plate image of the present invention.

[0028] Figure 7 It is a network architecture diagram of the license plate detection algorithm of the present invention.

[0029] Figure 8 yes Figure 7 The prediction layer in the network architecture.

[0030] Fig. 9 It is a network architecture diagram of the license plate recognition algorithm of the present invention.

[0031] Fig.10 yes Fig. 9 The basic block diagram in .

[0032] Fig.11 It is a three-dimensional schematic diagram of the three-button control switch of the present invention.

[0033]

Explanation of symbols

[0034] Locomotive 10

[0035] Locomotive Starting Battery 101

[0036] Accommodation compartment 102

[0037] Front 103

[0038] Locomotive switch 104

[0039] Locomotive generator module 105

[0040] Protection device 106

[0041] Locomotive load 107

[0042] Locomotive voltage and current stabilizing device 108

[0043] Rear box 109

[0044] Handle 110

[0045] AI artificial intelligence device 20

[0046] Radar speed measurement module 201

[0047] Wireless network AP device 202

[0048] Temperature voltage detection device 203

[0049] Speaker / Buzzer 204

[0050] RFID active receiver 205

[0051] The first license plate image capturing device 30

[0052] The first image capture card device 301

[0053] The second license plate image capturing device 40

[0054] The second image capture card device 401

[0055] Display device 50

[0056] Image A / D PCB board / touch control board 501

[0057] Heat dissipation device 60

[0058] Power supply management device 70

[0059] Reset switch 701

[0060] Charging selection switch 702

[0061] Bluetooth communication charger 703

[0062] Three button control switch 80

[0063] First push button switch 801

[0064] Second push button switch 802

[0065] The third button switch 803

[0066] Step 1 S1

[0067] Step 2 S2

[0068] Step 3 S3

[0069] Step 4 S4 DETAILED DESCRIPTION

[0070] See also Figure 3 and Figure 1As shown, they are respectively a three-dimensional appearance schematic diagram of the present invention and a control system block schematic diagram of the first embodiment of the present invention. As shown in the figure: the present invention is a control system of an AI artificial intelligence motorcycle, which includes a motorcycle 10 provided with a motorcycle starting battery 101, an AI artificial intelligence device 20 electrically connected to the motorcycle starting battery 101, a first license plate image capture device 30 electrically connected to the AI ​​artificial intelligence device 20, a second license plate image capture device 40 electrically connected to the AI ​​artificial intelligence device 20, a display device 50 electrically connected to the AI ​​artificial intelligence device 20, a heat dissipation device 60 electrically connected to the AI ​​artificial intelligence device 20, and a power supply management device 70 electrically connected to the AI ​​artificial intelligence device 20.

[0071] When configured, the AI ​​device 20 and the heat dissipation device 60 may be disposed in the storage compartment 102 below the seat cushion of the motorcycle 10, the first license plate image capture device 30 and the second license plate image capture device 40 may be installed on the front 103 of the motorcycle 10, the display device 50 may be disposed above the dashboard of the front 103, and the power supply management device 70 may be disposed in the rear box 109 of the motorcycle 10, and may be integrated with the AI ​​device 20, or may be disposed in another appropriate position of the motorcycle 10; in this way, The AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50, the heat dissipation device 60 and the power supply management device 70 can be installed on an existing motorcycle 10 to at least control the license plate image capture, license plate detection, license plate recognition calculation control and power supply management control; thus, when patrolling on the motorcycle 10, the effects of license plate recognition and active detection are achieved, thereby achieving better detection and prevention mobility and detection and prevention timeliness.

[0072] like Figure 1 As shown, the power supply management device 70 at least performs the step of turning on the AI ​​artificial intelligence device 20, the step of turning off the AI ​​artificial intelligence device 20, and the step of charging the locomotive starting battery 101.

[0073] When executing the startup step of the AI ​​artificial intelligence device 20, the locomotive power switch 104 provided on the locomotive starting battery 101 is turned on, and the reset switch 701 of the power supply management device 70 is turned on to turn on the locomotive starting battery 101 to output power to the AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50 and the heat dissipation device 60, and the display device 50 displays the startup procedure and the system screen to complete the startup of the AI ​​artificial intelligence device 20.

[0074] When executing the shutdown step of the AI ​​artificial intelligence device 20, the locomotive power switch 104 is turned off, and the AI ​​artificial intelligence device 20 enters the shutdown procedure. After the display device 50 displays that the shutdown is completed, the reset switch 701 is activated. When the display device 50 has no output image, the shutdown of the AI ​​artificial intelligence device 20 is completed.

[0075] When the charging step of the locomotive starting battery 101 is executed, the charging selection switch 702 of the power supply management device 70 is turned to the vehicle charging side position, so that the power supply management device 70 receives the power of the locomotive generator module 105 of the locomotive 10 to provide power to the input end of the power supply management device 70 for charging the locomotive starting battery 101, and can indirectly or directly supply power to the AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 10, and the like. The power required for the operation of the device 50 and the heat dissipation device 60; in addition, the charging selection switch 702 of the power supply management device 70 can also be activated to the external charging side position, so that the Bluetooth communication charger 703 of the power supply management device 70 receives AC power from the external power supply to charge the motorcycle starting battery 101, so as to indirectly or directly supply the power required for the operation of the AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50 and the heat dissipation device 60.

[0076] Based on the above operation mode, the charging selection switch 702 can charge the AI ​​artificial intelligence device 20 when the kinetic energy of the motorcycle 10 is converted into electrical energy while the motorcycle is moving; or when the motorcycle is parked, the charging selection switch 702 selects the external charging mode, and the power supply management device 70 receives the external power supply voltage for charging; and the Bluetooth communication charger 703 can be connected point-to-point with the mobile phone / computer device, and the charging can be controlled through the exclusive App software, and the communication distance can reach 15 meters.

[0077] In addition to the above embodiments, in the first embodiment of the present invention, the locomotive power switch 104 is further electrically connected to at least one protection device 106, a locomotive load 107, and a locomotive voltage and current stabilizing device 108, and the locomotive generator module 105 is electrically connected to the locomotive voltage and current stabilizing device 108. Thus, when actually used in a locomotive 10 with an existing fuel system, the protection device 106, the locomotive load 107 and the settings of the locomotive voltage and current stabilizing device 108 can be utilized to enable the locomotive 10 to be retrofitted with the AI ​​artificial intelligence device 20, and the three coil windings of the generator on the existing locomotive 10 can be enlarged to make the output current reach 8A to 10A (or replace another set of generators that meet the output power); the rectifier and voltage stabilizing device on the original locomotive 10 is replaced to meet the system requirements; and a protection device 106 (such as a power diode) is installed on the load side of the original locomotive 10 to reduce the increased current value to the original locomotive power supply value. At the same time, the increased output current can power the entire AI system for operation for more than 8 hours, achieving oil-electricity symbiosis.

[0078] Based on the above-mentioned implementation, the existing locomotive 10 can be used to add the AI ​​artificial intelligence device 20 to change the locomotive generator module 105 on the locomotive 10. The kinetic energy of the movement drives the locomotive generator module 105, and a super capacitor is installed as needed. The output reaches DC12V / 15A through the locomotive voltage and current stabilizing device 108 to supply power to the power supply management device 70 for charging, thereby indirectly or directly supplying the power required for the operation of the AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50 and the heat dissipation device 60.

[0079] In addition to the above embodiments, in the first embodiment of the present invention, the AI ​​artificial intelligence device 20 is further electrically connected to a radar speed measurement module 201 and a wireless network AP device 202. The wireless network AP device 202 serves as a wireless connection between the AI ​​artificial intelligence device 20 and other external devices. The wireless network AP device 202 can be a WAP device or a wireless network base station, and has a bridging function. It can be installed inside the accommodation compartment 102 of the locomotive 10 according to the actual requirements of the locomotive 10, or installed inside the front 103 of the locomotive 10, so that the AI ​​artificial intelligence device 20 can achieve the effect of wireless connection with other external devices.

[0080] In addition to the above embodiments, in the first embodiment of the present invention, the AI ​​artificial intelligence device 20 is electrically connected to a temperature and voltage detection device 203, and the temperature and voltage detection device 203 can communicate with the AI ​​artificial intelligence device 20 in a two-way manner through a USB port and interface software. When the temperature and voltage detection device 203 detects the temperature / voltage / current values ​​of the first license plate image capture device 30, the second license plate image capture device 40, and the display device 50, and compares them with the safety values ​​set by the AI ​​artificial intelligence device 20, the heat dissipation device 60 is driven to inlet and exhaust air so that the accommodating compartment 102 forms a strong convection airflow to reduce the temperature.

[0081] Based on the above embodiment, the power supply management device 70 is a power supply device specially developed for the system of the locomotive 10, and is connected to the AI ​​artificial intelligence device 20 to serve as a system voltage and current stabilization power supply, power input and output control, and digital power management function; and has an emergency rescue function if the original battery on the locomotive 10 fails and cannot start the vehicle.

[0082] Furthermore, the AI ​​device 20 is provided with another ACC function connection, which is connected to the locomotive switch 104 of the locomotive 10, so as to detect the DC12V conduction identification when the locomotive 10 is started, and initiate the startup action of the AI ​​device 20; and whether the AI ​​device 20 is started will depend on the opening of the output end of the power supply management device 70. In the same situation, when the locomotive switch 104 of the locomotive 10 is turned off, the ACC connection has no voltage input identification, and the AI ​​device 20 is turned off; and the AI ​​device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50 and the heat dissipation device 60 are disconnected from the power supply, which will be regarded as the shutdown of the output end of the power supply management device 70 to avoid the AI ​​device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50 and the heat dissipation device 60 being connected for a long time, resulting in a trickle current consumption situation.

[0083] See also Figure 3 and Figure 2 As shown in the figure, they are respectively a three-dimensional appearance schematic diagram of the present invention and a control system block schematic diagram of the second embodiment of the present invention. As shown in the figure: the present invention is a control system for an AI artificial intelligence locomotive, which includes:

[0084] A locomotive 10, such as an existing locomotive 10; an AI artificial intelligence device 20, such as an AI computer, which is arranged in a storage compartment 102 of the locomotive 10; a first license plate image capture device 30, which is electrically connected to the AI ​​artificial intelligence device 20, and the first license plate image capture device 30 includes a first image capture card device 301, and the first image capture card device 30 is installed on the front 103 of the locomotive 10, and the first image capture card device 301 is electrically connected to the first license plate image capture device 30 through a coaxial cable transmission signal through an analog image signal input end, so that the digital image signal output end of the first license plate image capture device 30 is electrically connected to the AI ​​artificial intelligence device 20 through a wire transmission signal; this wiring layout not only considers the appropriateness of the equipment installation environment factors, but also considers avoiding the analog image signal from being electromagnetically interfered by the motor or inductive load device of the locomotive 10 transmission system.

[0085] A second license plate image capture device 40 is electrically connected to the AI ​​artificial intelligence device 20. The second license plate image capture device 40 includes a second image capture card device 401. The second image capture card device 401 is installed on the front of the motorcycle 103. The second image capture card device 401 is electrically connected to the second license plate image capture device 40 through a coaxial cable to transmit a signal at an analog image signal input end, so that the digital image signal output end of the second license plate image capture device 40 is electrically connected to the AI ​​artificial intelligence device 20 through a wire transmission signal. In addition to considering the appropriateness of the equipment installation environment, this wiring layout also takes into account avoiding electromagnetic interference of the analog image signal from the motor or inductive load device of the motorcycle 10 transmission system.

[0086] A display device 50 is electrically connected to the AI ​​artificial intelligence device 20. The display device 50 includes an image A / D PCB board / touch control board 501. The display device 50 is arranged above the dashboard of the front of the vehicle 103. The image A / D PCB board / touch control board 501 is arranged in the accommodation compartment 102. The image A / D PCB board / touch control board 501 is electrically connected to the display device 50 by a bonding line. The image A / D PCB board / touch control board 501 is moved out of the screen body and arranged in the accommodation compartment 102 by simplifying the bonding line and strengthening the tensile strength of the wire. This design is to consider the adverse environmental factors of the external touch screen, which are exposed to wind, rain, and high temperature in the sun for a long time, and produce internal moisture and high temperature, which affect the long-term durable operation function of the electronic parts; and the use consideration that the components may be subjected to excessive vibration due to uneven roads, resulting in the breakage of the fixing point or the shear force of the substrate. Furthermore, the components are configured in a separate manner, which is convenient for future maintenance.

[0087] A wireless network AP device (router) 202 is electrically connected to the AI ​​artificial intelligence device 20 through an RJ45 port network cable, as a point-to-point wireless connection between the AI ​​artificial intelligence device 20 and external devices, and has the function of gain-amplifying the communication range. The network communication covers an unobstructed range of more than 30 meters; or a 5G Wifi network module is directly built into the AI ​​artificial intelligence device 20 housing, wired to the computer housing antenna seat, and then led to the outside of the vehicle body as a point-to-point wireless connection between the AI ​​artificial intelligence device 20 and external devices. And the external device (peripheral device) can be a smart phone or smart glasses as needed. In terms of application, the App installed on the smart phone or smart glasses starts a two-way communication with the AI ​​artificial intelligence device 20, and captures the real-life image taken by the camera on the mobile phone or glasses, and transmits it to the AI ​​artificial intelligence device 20 through WebRTC software or other self-opening mobile phone image wireless communication mechanism for image calculation analysis, and then sends the analysis result information back to the smart phone or smart glasses to meet the needs of actual application.

[0088] A radar speed measurement module 201 is electrically connected to the AI ​​device 20 and is disposed in the rear box 109 of the motorcycle 10. The radar speed measurement module 201 is connected to the RS485 port of the AI ​​device 20 via a UPT signal line, or can be connected to the RS232 port of the AI ​​device 20 via an RS485 / RS232 converter, or can be connected to the USB port of the AI ​​device 20 via an RS485 / USB converter. When the motorcycle 10 is parked on the roadside, the radar speed measurement module 201 cooperates with the license plate recognition function of the license plate image capture devices 30, 40, and can instantly calculate the speed of each moving vehicle and the corresponding license plate number through an image analysis algorithm.

[0089] An RFID active receiver 205 is electrically connected to the AI ​​device 20 and is disposed on the front 103 of the motorcycle 10. The RFID active receiver 205 is connected to the RS485 port of the AI ​​device 20 via a UPT signal line, or can be connected to the RS232 port of the AI ​​device 20 via an RS485 / RS23 converter, or can be connected to the USB port of the AI ​​device 20 via an RS485 / USB converter. The RFID active receiver 205 can receive ETC / eTag chips on parking vehicles charged for roadside parking while the motorcycle 10 is moving, and cooperate with the license plate recognition function of the license plate image capture devices 30, 40. Through an image analysis algorithm, the speed of each moving vehicle and the corresponding license plate number can be calculated in real time.

[0090] Thus, the AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50, the radar speed measurement module 201 and the RFID active receiver 205 can be installed on the existing motorcycle 10, so as to at least control the license plate image capture, license plate detection, license plate recognition calculation control and power supply management control; thus, when patrolling on the motorcycle 10, the license plate recognition and active detection effects are achieved, thereby achieving better detection and prevention mobility and detection and prevention timeliness.

[0091] In addition to the above embodiments, in the first embodiment of the present invention, the AI ​​artificial intelligence device 20 is electrically connected to a speaker / buzzer 204, and the speaker / buzzer 204 is provided on the display device 50. The trigger / signal amplification function of the speaker / buzzer 204 is connected to the image A / D signal through the signal control cable. The PCB board / touch control board 501 has an embedded amplifier element connected to the actuator, and the speaker / buzzer 204 is connected to the GPIO / RS232 adapter of the AI ​​artificial intelligence device 20 through a UTP signal line, or through the embedded sound source signal conversion control board on the AI ​​artificial intelligence device 20, the analog sound source signal is converted into a level voltage drop signal, the speaker hole of the AI ​​artificial intelligence device 20 is connected to the buzzer, and the timing of the alarm mechanism is controlled by the AI ​​artificial intelligence device 20; in this way, when the motorcycle 10 is driving, when a related event occurs, the speaker / buzzer 204 can immediately send out a notification / warning sound, and connect the event alarm mechanism timing of the AI ​​artificial intelligence device 20 (for example, the license plate recognized during driving matches the database list or the system detection value exceeds / is lower than the set range value), so that the present invention can better meet the needs of practical application.

[0092] In addition to the above embodiments, in the first embodiment of the present invention, the digital image signal output end of the first license plate image capture device 30 is electrically connected to the AI ​​artificial intelligence device 20 via a USB3.0 cable transmission signal, and the digital image signal output end of the second license plate image capture device 40 is electrically connected to the AI ​​artificial intelligence device 20 via a USB3.0 cable transmission signal; this wiring layout not only considers the appropriateness of the equipment installation environment factors, but also considers avoiding the analog image signal from being electromagnetically interfered by the motor of the locomotive 10 transmission system or the inductive load device, so that the present invention can better meet the needs of actual application.

[0093] In addition to the above embodiments, Fig.11As shown, in the first and second embodiments of the present invention, the AI ​​artificial intelligence device 20 is electrically connected to a three-button control switch 80, which is disposed at the position of the handlebar 110 of the motorcycle 10 and is electrically connected to the AI ​​artificial intelligence device 20 through a control line. The three-button control switch 80 is a single-loop switching switch, and the control panel of the three-button control switch 80 is disposed in the AI ​​artificial intelligence device 20 with a USB port; the first button switch 801 of the three-button control switch 80 controls the temporary power off or power off of the display device 50 through the AI ​​artificial intelligence device 20; the second button switch 802 of the three-button control switch 80 activates a soft shutdown program action through the AI ​​artificial intelligence device 20 to shut down; the third button switch 803 of the three-button control switch 80 controls whether the image stream code of the first license plate image capture device 30 enters the recognition program and whether the image stream code of the second license plate image capture device 40 enters the recognition program through the built-in software of the AI ​​artificial intelligence device 20; thereby, the present invention can better meet the needs of practical application.

[0094] See also Figure 3 and Figure 4 As shown, they are respectively a schematic diagram of the three-dimensional appearance of the present invention and a schematic diagram of the license plate recognition control system process of the third embodiment of the present invention. As shown in the figure: the present invention is a license plate recognition control system for an AI artificial intelligence motorcycle, which includes:

[0095] A locomotive 10 is provided with a locomotive starting battery 101, an AI artificial intelligence device 20 electrically connected to the locomotive starting battery 101, a first license plate image capturing device 30 electrically connected to the AI ​​artificial intelligence device 20, a second license plate image capturing device 40 electrically connected to the AI ​​artificial intelligence device 20, a display device 50 electrically connected to the AI ​​artificial intelligence device 20, a heat dissipation device 60 electrically connected to the AI ​​artificial intelligence device 20, and a power supply management device 70 electrically connected to the AI ​​artificial intelligence device 20.

[0096] When configured, the AI ​​device 20 and the heat dissipation device 60 may be disposed in the accommodation compartment 102 of the motorcycle 10, the first license plate image capture device 30 and the second license plate image capture device 40 may be installed on the front 103 of the motorcycle 10, the display device 50 may be disposed above the dashboard of the front 103, and the power supply management device 70 may be disposed in the rear box 109 of the motorcycle 10, and may be integrated with the AI ​​device 20, or may be disposed in another appropriate position of the motorcycle 10; in this way, The AI ​​artificial intelligence device 20, the first license plate image capture device 30, the second license plate image capture device 40, the display device 50, the heat dissipation device 60 and the power supply management device 70 can be installed on an existing motorcycle 10 to control the license plate image capture, the calculation control of license plate detection and license plate recognition, and the control of power supply management; thus, when patrolling on the motorcycle 10, the effects of license plate recognition and active detection are achieved, thereby achieving better detection and prevention mobility and detection and prevention timeliness.

[0097] See also Figure 4 and Figure 5 As shown, the AI ​​artificial intelligence device 20 cooperates with the first license plate image capturing device 30 and the second license plate image capturing device 40 to perform a license plate recognition operation; and the license plate recognition operation includes at least the following steps:

[0098] Step 1 S1: Scene Image Input, first obtain at least one license plate image from the first license plate image capturing device 30 and the second license plate image capturing device 40 and input it into the AI ​​artificial intelligence device 20 .

[0099] Step 2 S2: License Plate Detection, based on a neural network-based license plate detection algorithm, detects the license plate frame position, size, and tilt angle of the characters on the license plate in the license plate image, and cuts the detected license plate out of the license plate image.

[0100] Step 3 S3: Rotation Correction: according to the inclination angle and classification of the characters on the license plate determined in step 2, the license plate in the license plate image is subjected to rotation correction so that the characters on the license plate do not overlap as much as possible in the horizontal straight line position.

[0101] Step 4 S4: Plate Number Recognition: input the rotation-corrected license plate image into a license plate recognition neural network to recognize the license plate number.

[0102] Please refer to Figures 5 to 10. Figure 5 is a flow chart of the license plate recognition algorithm of the present invention, Figure 6is a classification diagram of the inclination angle of the license plate image of the present invention, Figure 7 is a network architecture diagram of the license plate detection algorithm of the present invention, Figure 8 for Figure 7 The prediction layer in the network architecture of Fig. 9 is a network architecture diagram of the license plate recognition algorithm of the present invention, Fig.10 for Fig. 9 The basic blocks in .

[0103] In addition to the above embodiments, Figure 7 and Figure 8 As shown, in the third embodiment of the present invention, the depthwise separable convolution of the neural network-like license plate recognition algorithm in step two S2 gradually extracts the features of the license plate in the license plate image from 512 by 512 to a smaller size, and performs prediction through prediction layers in six depths; in this way, the AI ​​artificial intelligence device 20 can be used to cooperate with the first license plate image capture device 30 and the second license plate image capture device 40 to perform license plate detection operations.

[0104] In addition to the above embodiments, in the third embodiment of the present invention, the size of the first prediction layer (PredictionLager) is 32 times 32, the second prediction layer is 16 times 16, the third prediction layer is 8 times 8, the fourth prediction layer is 4 times 4, the fifth prediction layer is 2 times 2, and the sixth prediction layer is 1 times 1. After obtaining the license plate position and tilt angle classification in the license plate image, it is cut from the scene according to the coordinates and rotation correction is performed to complete the preprocessing of license plate recognition; thereby, the AI ​​artificial intelligence device 20 can be used to cooperate with the first license plate image capture device 30 and the second license plate image capture device 40 to perform license plate detection operations.

[0105] In addition to the above embodiments, Figure 6 As shown, in the third embodiment of the present invention, the license plate inclination angle in step two S2 is divided into 8 categories, namely two groups of angles of 0 to 10 degrees, two groups of angles of 10 to 20 degrees, two groups of angles of 20 to 30 degrees, and two groups of angles of more than 30 degrees, and the detected license plate is cut out from the license plate image; in this way, the AI ​​artificial intelligence device 20 can be used to cooperate with the first license plate image capture device 30 and the second license plate image capture device 40 to perform license plate detection operations.

[0106] like Fig. 9 and Fig.10As shown, in the third embodiment of the present invention, after the above-mentioned preprocessing, the license plate image has been obtained, and then the license plate recognition network of step four S4 is used to identify the license plate string result. The license plate image is fixed to a rectangular image of 48*94 pixels and enters the license plate recognition network. After four convolutional layers, three maximum pooling layers, and one fully connected layer, the features of the license plate will be transformed into a string of 1*1 points according to the vertical direction. According to the characteristics of the vertical position, the license plate recognition network will output a probability map for each 1*1 feature. According to the probability map, a predicted string can be obtained. Finally, the repeated words are merged through CTC (Connectionist Temporal Classification), and the blank words are removed to obtain the final license plate number result.

[0107] In summary, the control system of the AI ​​artificial intelligence motorcycle of the present invention effectively improves the conventional deficiencies. The AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device, the heat dissipation device and the power supply management device can be mounted on an existing motorcycle to control the license plate image capture, the license plate detection, the calculation control of the license plate recognition and the power supply management. Therefore, when the motorcycle 10 is used for patrolling, the license plate recognition and active detection effects are achieved, thereby achieving better detection and prevention mobility and detection and prevention instantaneity. The present invention has the advantages of progress and practicality and meets the needs of users. It has the basic requirements for patent application and can be applied for according to law.

[0108] The above is only a preferred embodiment of the present invention and should not be used to limit the technical scope of the present invention. Therefore, any simple equivalent changes and modifications made with reference to the claims and description of the present invention should still fall within the scope of the patent to be obtained by the present invention in the future.

Claims

1. A control system for an AI artificial intelligence locomotive, characterized in that: include: A motorcycle provided with a motorcycle starting battery, an AI artificial intelligence device electrically connected to the motorcycle starting battery, a first license plate image capturing device electrically connected to the AI ​​artificial intelligence device, a second license plate image capturing device electrically connected to the AI ​​artificial intelligence device, a display device electrically connected to the AI ​​artificial intelligence device, a heat dissipation device electrically connected to the AI ​​artificial intelligence device, and a power supply management device electrically connected to the AI ​​artificial intelligence device; The AI ​​artificial intelligence device and the heat dissipation device are disposed in the accommodation compartment of the motorcycle, the first license plate image capture device and the second license plate image capture device are installed on the front of the motorcycle, and the display device is disposed above the dashboard of the front of the motorcycle; The power supply management device at least performs a step of turning on the AI ​​artificial intelligence device, a step of turning off the AI ​​artificial intelligence device, and a step of charging the battery for starting the locomotive; When executing the step of starting the AI ​​artificial intelligence device, the locomotive power switch provided for the locomotive starting battery is turned on, and the reset switch of the power supply management device is turned on to turn on the locomotive starting battery to output power to the AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device, and the display device displays the boot procedure and system screen to complete the startup of the AI ​​artificial intelligence device; When executing the shutdown step of the AI ​​device, turn off the locomotive power switch, and the AI ​​device enters the shutdown procedure. When the display device displays that the shutdown is completed, start the reset switch, and when the display device has no output image, the shutdown of the AI ​​device is completed. When executing the charging step of the locomotive starting battery, the charging selection switch of the power supply management device is activated to the vehicle charging side position, so that the power supply management device receives the electric energy of the locomotive generator module of the locomotive to provide power to the input end of the power supply management device for charging the locomotive starting battery, and indirectly or directly supplies the power required for the operation of the AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device; in addition, the charging selection switch of the power supply management device can also be activated to the external charging side position, so that the Bluetooth communication charger of the power supply management device receives AC power from an external power supply for charging the locomotive starting battery, and indirectly or directly supplies the power required for the operation of the AI ​​artificial intelligence device, the first license plate image capture device, the second license plate image capture device, the display device and the heat dissipation device.

2. The control system of the AI ​​artificial intelligence locomotive according to claim 1, characterized in that: The locomotive switch is further electrically connected to at least one protection device, a locomotive load, and a locomotive voltage and current stabilizing device, and the locomotive generator module is electrically connected to the locomotive voltage and current stabilizing device.

3. The control system of the AI ​​artificial intelligence locomotive according to claim 1, characterized in that: The AI ​​device is further electrically connected to a radar speed measurement module and a wireless network AP device or the AI ​​device has a built-in 5G Wifi module.

4. The control system of the AI ​​artificial intelligence locomotive according to claim 3, characterized in that: The wireless network AP device serves as a wireless connection between the AI ​​device and other external devices. The external device can be a smartphone or smart glasses as needed. An App installed on the smartphone or smart glasses can initiate two-way communication with the AI ​​device and capture real-life images taken by the camera on the smartphone or glasses. The images are transmitted to the AI ​​device through WebRTC software or other self-activated mobile phone image wireless communication mechanisms for image calculation and analysis, and the analysis result information is then transmitted back to the smartphone or smart glasses.

5. The control system of the AI ​​artificial intelligence locomotive according to claim 1, characterized in that: The AI ​​artificial intelligence device is electrically connected to a temperature and voltage detection device, and the temperature and voltage detection device can communicate with the AI ​​artificial intelligence device in a two-way manner through a USB port or Bluetooth wireless communication and interface software. The temperature and voltage detection device detects the temperature / voltage / current values ​​of the first license plate image capture device, the second license plate image capture device, and the display device, and compares them with the safety values ​​set by the AI ​​artificial intelligence device, thereby driving the heat dissipation device to inlet and exhaust air, so that the accommodation compartment forms a strong convection airflow to reduce the temperature.

6. A control system for an AI locomotive, characterized in that: include: a locomotive; An AI artificial intelligence device, which is installed in the accommodation compartment of the locomotive; a first license plate image capture device, which is electrically connected to the AI ​​artificial intelligence device, wherein the first license plate image capture device comprises a first image capture card device, which is installed on the front of the motorcycle, and the first image capture card device is electrically connected to the first license plate image capture device through a coaxial cable transmission signal at an analog image signal input end, so that the digital image signal output end of the first license plate image capture device is electrically connected to the AI ​​artificial intelligence device through a wire transmission signal; a second license plate image capture device, which is electrically connected to the AI ​​artificial intelligence device, the second license plate image capture device includes a second image capture card device, the second image capture card device is installed on the front of the motorcycle, and the second image capture card device is electrically connected to the second license plate image capture device through a coaxial cable transmission signal with an analog image signal input end, so that the digital image signal output end of the second license plate image capture device is electrically connected to the AI ​​artificial intelligence device through a wire transmission signal; a display device, which is electrically connected to the AI ​​artificial intelligence device, the display device includes an image A / D PCB board / touch control board, the display device is arranged above the dashboard of the front of the vehicle, the image A / D PCB board / touch control board is arranged in the accommodating compartment, and the image A / D PCB board / touch control board is electrically connected to the display device by a bonding line; A wireless network AP device, which is electrically connected to the AI ​​device via an RJ45 port network cable or a USB port, or the AI ​​device has a built-in 5G Wifi module as a bridge to connect the AI ​​device to external devices or a cloud server wirelessly; a radar speed measurement module, which is electrically connected to the AI ​​device and is disposed in the rear box of the motorcycle. The radar speed measurement module is connected to the RS485 port of the AI ​​device via a UPT signal line, or can be connected to the RS232 port of the AI ​​device via an RS485 / RS232 converter, or can be connected to the USB port of the AI ​​device via an RS485 / USB converter. When the motorcycle is parked on the roadside, the radar speed measurement module cooperates with the license plate recognition function of the first license plate image capture device and the second license plate image capture device, and can instantly calculate the speed of each moving vehicle and the matching license plate number through an image analysis algorithm; An RFID active receiver is electrically connected to the AI ​​device and is disposed on the front of the motorcycle. The RFID active receiver is connected to the RS485 port of the AI ​​device via a UPT signal line, or can be connected to the RS232 port of the AI ​​device via an RS485 / RS23 converter, or can be connected to the USB port of the AI ​​device via an RS485 / USB converter. The RFID active receiver can receive ETC / eTag chips on parking vehicles charged for roadside parking while the motorcycle is moving, and cooperate with the license plate recognition functions of the first license plate image capture device and the second license plate image capture device. Through an image analysis algorithm, the speed of each moving vehicle and the corresponding license plate number can be calculated in real time.

7. The control system of the AI ​​artificial intelligence locomotive according to claim 6, characterized in that: The AI ​​device is further electrically connected to a speaker / buzzer, which is provided on the display device. The triggering / signal amplification function of the speaker / buzzer is activated by connecting the signal control cable to the amplifier component embedded in the image A / D PCB board / touch control board, and the speaker / buzzer is connected to the GPIO / RS232 adapter of the AI ​​device through a UTP signal line, or through the embedded sound source signal conversion control board on the AI ​​device to convert the analog sound source signal into a level voltage drop signal. The speaker hole of the AI ​​device is connected to the buzzer, and the timing of the reporting mechanism is controlled by the AI ​​device.

8. The control system of the AI ​​artificial intelligence locomotive according to claim 6, characterized in that: The digital image signal output end of the first license plate image capture device is electrically connected to the AI ​​artificial intelligence device via a USB3.0 cable transmission signal, and the digital image signal output end of the second license plate image capture device is electrically connected to the AI ​​artificial intelligence device via a USB3.0 cable transmission signal.

9. The control system of the AI ​​artificial intelligence locomotive according to claim 6, characterized in that: The AI ​​device is electrically connected to a three-button control switch, which is disposed on the handle of the motorcycle and is electrically connected to the AI ​​device through a control line. The three-button control switch is a single-loop switching switch, and a control panel of the three-button control switch is disposed in the AI ​​device with a USB port. The first button switch of the three-button control switch controls the temporary power supply or power supply cutoff of the display device through the AI ​​device; the second button switch of the three-button control switch starts a slow shutdown program action through the AI ​​device to shut down; the third button switch of the three-button control switch controls whether the image stream code of the first license plate image capture device enters the recognition program and whether the image stream code of the second license plate image capture device enters the recognition program through the built-in software of the AI ​​device.

10. A license plate recognition control system for an AI artificial intelligence locomotive, characterized in that: include: A motorcycle provided with a motorcycle starting battery, an AI artificial intelligence device electrically connected to the motorcycle starting battery, a first license plate image capturing device electrically connected to the AI ​​artificial intelligence device, and a second license plate image capturing device electrically connected to the AI ​​artificial intelligence device, wherein the AI ​​artificial intelligence device is disposed in a storage compartment of the motorcycle, and the first license plate image capturing device and the second license plate image capturing device are installed on the front of the motorcycle; The AI ​​artificial intelligence device cooperates with the first license plate image capturing device and the second license plate image capturing device to perform license plate recognition calculation; The operation of the license plate recognition includes the following steps: Step 1: License plate image input, first obtain at least one license plate image from the first license plate image capture device and the second license plate image capture device and input it into the AI ​​artificial intelligence device; Step 2: License plate detection, based on a neural network-like license plate recognition algorithm, detecting the license plate frame position, size and tilt angle of the characters on the license plate in the license plate image, and cutting the detected license plate out of the license plate image; Step 3: rotation correction, according to the classification of the tilt angle of the characters on the license plate determined in step 2, the license plate in the license plate image is rotated and corrected so that the characters on the license plate do not overlap as much as possible in the horizontal straight line position; Step 4: License plate number recognition: input the rotation-corrected license plate image into the license plate recognition neural network to recognize the license plate number.

11. The license plate recognition control system for an AI artificial intelligence motorcycle according to claim 10, characterized in that: The depthwise separable convolution of the neural network-like license plate recognition algorithm in step 2 gradually extracts the features of the license plate in the license plate image from 512 by 512 to a smaller size, and makes predictions through prediction layers in six depths.

12. The license plate recognition control system for an AI artificial intelligence motorcycle according to claim 11, characterized in that: The size of the first prediction layer is 32 times 32, the second prediction layer is 16 times 16, the third prediction layer is 8 times 8, the fourth prediction layer is 4 times 4, the fifth prediction layer is 2 times 2, and the sixth prediction layer is 1 times 1. After obtaining the license plate position and angle classification in the license plate image, it is cut from the scene according to the coordinates and rotation correction is performed to complete the preprocessing of license plate recognition.

13. The license plate recognition control system for an AI artificial intelligence motorcycle according to claim 10, characterized in that: The tilt angles in step three are divided into 8 categories, namely, two groups of left and right 0 to 10 degrees, two groups of left and right 10 to 20 degrees, two groups of left and right 20 to 30 degrees, and two groups of angles above 30 degrees, and the detected license plate is cut out from the license plate image.

14. The license plate recognition control system for an AI artificial intelligence motorcycle according to claim 10, characterized in that: The license plate number recognition in step 4 uses the preprocessed license plate image and the license plate recognition network to identify the license plate string result.

Images