CN214751919U - Traffic sign recognition and detection device based on deep convolutional neural network - Google Patents
Traffic sign recognition and detection device based on deep convolutional neural network Download PDFInfo
- Publication number
- CN214751919U CN214751919U CN202120759449.1U CN202120759449U CN214751919U CN 214751919 U CN214751919 U CN 214751919U CN 202120759449 U CN202120759449 U CN 202120759449U CN 214751919 U CN214751919 U CN 214751919U
- Authority
- CN
- China
- Prior art keywords
- neural network
- traffic sign
- convolutional neural
- deep convolutional
- insulating shell
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 34
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 31
- 240000007651 Rubus glaucus Species 0.000 claims abstract description 31
- 235000011034 Rubus glaucus Nutrition 0.000 claims abstract description 31
- 235000009122 Rubus idaeus Nutrition 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 20
- 230000008093 supporting effect Effects 0.000 claims description 39
- 230000007246 mechanism Effects 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 abstract description 6
- 238000009434 installation Methods 0.000 abstract description 3
- 230000009977 dual effect Effects 0.000 abstract 1
- 238000013461 design Methods 0.000 description 3
- 238000000034 method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000003014 reinforcing effect Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The utility model discloses a traffic sign discernment detection device based on degree of depth convolution neural network belongs to motor vehicle driver assistance tool field. The system comprises an image acquisition unit and a processing unit; the processing unit comprises an insulating shell; the sucker seat is detachably connected with the interior of the motor vehicle, and the top of the sucker seat is connected with the bottom of the insulating shell; the display screen is embedded in one end face of the insulating shell; a speaker fixed to one side surface of the insulating housing; the raspberry pie is fixed inside the insulating shell, an input interface is connected with the image acquisition unit through a wire, and an output interface of the raspberry pie is connected with the display screen and the loudspeaker through a wire; the raspberry pi is a raspberry pi carrying a deep convolutional neural network algorithm. The utility model discloses simple structure, the installation is dismantled all very conveniently, can pass through the sense of hearing, the more comprehensive presentation of vision dual mode to the detection identification result of traffic sign for the driver, and the recognition rate is fast, detects the precision height, uses the scene extensive.
Description
Technical Field
The utility model belongs to motor vehicle driver assistance tool field, more specifically say, relate to a traffic sign discernment detection device based on degree of depth convolution neural network.
Background
Along with the rapid increase of the number of private cars, frequent traffic accidents, overspeed driving, fatigue driving, driving without a traffic sign and the like are main reasons causing car accidents, so the importance of equipment which can be automatically identified and alarmed through the traffic sign is gradually reflected on the motor vehicles, and in addition, sensing the surrounding environment and accurately identifying the traffic sign are also an important component of the automatically driven cars.
At present, the research aiming at intelligent recognition of road traffic signs is carried out at home and abroad, various recognition algorithms are provided starting from the aspects of digital image processing, traditional machine learning, deep neural network and the like, but the practical requirements can not be met from the aspects of stability and simplicity, and the problems of low recognition rate, high omission factor, long algorithm training time and the like exist; when data are changed into a complete natural scene road image, the traditional thought is difficult to obtain a reliable result, and due to the complexity of road traffic, the traffic sign identification system still has reliable accuracy under the influences of factors such as illumination, weather, sign aging, background interference, environmental shielding and the like under various complex scenes; the existing vehicle-mounted device for identifying and detecting the traffic sign is usually high in cost, a plurality of structural parts need to be additionally arranged and modified on a motor vehicle, and a user needs to pay a large cost for transferring the traffic identification and detection device after replacing the motor vehicle.
Disclosure of Invention
In order to solve the above problems, the utility model adopts the following technical proposal.
A traffic sign recognition and detection device based on a deep convolutional neural network comprises,
the image acquisition unit is arranged on the front side of the motor vehicle, and is used for acquiring and transmitting image information in front of the motor vehicle;
the processing unit is arranged in the motor vehicle and is in signal connection with the image information acquisition unit;
the processing unit comprises a processing unit and a control unit,
an insulating housing;
the sucker seat is detachably connected with the interior of the motor vehicle, and the top of the sucker seat is connected with the bottom of the insulating shell;
the display screen is embedded in one end face of the insulating shell;
a speaker fixed to one side surface of the insulating housing;
the raspberry pie is fixed inside the insulating shell, an input interface of the raspberry pie is connected with the image acquisition unit through a wire, and an output interface of the raspberry pie is respectively connected with the display screen and the loudspeaker through a wire;
the raspberry pie is a raspberry pie loaded with a deep convolutional neural network algorithm.
Further, the image acquisition unit includes:
a base fixed to a front side of the motor vehicle;
the camera is arranged above the base and used for collecting image information in front of the motor vehicle, and a polaroid is arranged on the camera lens.
Further, the image acquisition unit further includes:
the rotating support is fixed on the top surface of the base;
and one end of the adjustable hose is fixedly connected with the top of the rotary support, and the other end of the adjustable hose is fixedly connected with the camera.
Furthermore, a position giving opening is formed in the other end face, opposite to the display screen, of the insulating shell;
still include adjustable supporting mechanism, it includes:
the adjusting piece is arranged in the insulating shell and comprises a transverse plate and two side plates, the transverse plate is horizontally arranged, the two side plates are parallel to each other and are respectively fixed at two ends of the transverse plate, one side of each side plate is fixed with the end part of the transverse plate, and a tooth-shaped bulge is formed at the bottom of the other side of each side plate;
the supporting piece is arranged below the adjusting piece and corresponds to the position of the abdicating opening, the supporting piece comprises two supporting rods which are respectively arranged below the two side plates correspondingly, one side of the top surface of each supporting rod is vertically provided with a connecting plate, the connecting plate is rotatably connected with the side plates, and the other side of the top surface of each supporting rod is provided with a tooth-shaped recess which is matched and connected with the tooth-shaped protrusion.
Furthermore, a guide rail is arranged on the inner wall above the abdicating opening of the insulating shell, the adjusting piece is connected in the guide rail in a sliding manner, and a lifting handle opening is formed in the top surface of the insulating shell right above the guide rail;
the lifting handle is fixed on the tops of the two connecting plates and is positioned right below the lifting handle opening.
Furthermore, at least one of the tooth-shaped protrusion and the tooth-shaped recess is of an elastic structure.
Advantageous effects
Compared with the prior art, the beneficial effects of the utility model are that:
(1) the utility model discloses a traffic sign discernment detection device based on degree of depth convolution neural network, the installation is dismantled all very conveniently, and at the present that the iteration is changed comparatively frequently to the private car, the transfer cost of this device is very low, and this device can show the driver to the detection recognition result of traffic sign more comprehensively through sense of hearing, vision two kinds of modes, and this device adopts the raspberry group that carries the degree of depth convolution neural network algorithm to send out the detection and identification of image, and recognition speed is fast, and detection precision is high, the application scene is extensive;
(2) the utility model discloses a traffic sign discernment detection device based on degree of depth convolution neural network, the camera of image acquisition unit can adjust the angular position in a flexible way to the better acquisition image information that comes of adaptation different motorcycle types, driving environment;
(3) the traffic sign recognition and detection device based on the deep convolutional neural network has the advantages that the angle of the insulating shell is adjustable, and a display screen can be conveniently observed by a user at a proper angle;
(4) the traffic sign recognition and detection device based on the deep convolutional neural network is provided with the adjustable supporting mechanism, and can provide reliable position support for the insulation shell with a proper angle;
(5) the traffic sign recognition and detection device based on the deep convolutional neural network is also provided with a handle structure, so that the portability of the device is improved;
(6) the utility model discloses simple structure, reasonable in design easily makes.
Drawings
Fig. 1 is a schematic structural diagram of the traffic sign recognition and detection device based on the deep convolutional neural network of the present invention;
fig. 2 is a schematic view of the internal structure of the insulating housing of the present invention;
fig. 3 is another view angle structure diagram of the insulating housing of the present invention;
FIG. 4 is a schematic structural view of the adjustable supporting mechanism of the present invention;
FIG. 5 is a schematic structural view of the adjusting member of the present invention;
fig. 6 is a schematic view of another perspective structure of the adjusting member of the present invention;
FIG. 7 is a schematic structural view of the supporting member of the present invention;
fig. 8 is a schematic view of another perspective structure of the supporting member of the present invention;
fig. 9 is a flowchart of the traffic sign recognition and detection device based on the deep convolutional neural network according to the present invention;
in the figure:
1. an image acquisition unit; 10. a base; 11. a camera; 110. a polarizing plate; 12. rotating the support; 13. an adjustable hose;
2. a processing unit; 20. an insulating housing; 200. a let position port; 201. a handle opening; 202. a guide rail; 203. a limiting plate; 21. a sucker seat; 22. a display screen; 23. a speaker; 24. a raspberry pie; 25. an adjustable support mechanism; 250. an adjustment member; 2500. a transverse plate; 2501. a side plate; 2502. a tooth-shaped bulge; 251. a support member; 2510. a support bar; 2511. a connecting plate; 2512. tooth-shaped depressions; 2513. a reinforcing bar; 252. a handle.
Detailed Description
The invention will be further described with reference to specific embodiments and drawings.
Example 1
The traffic sign identification and detection device based on the deep convolutional neural network of the embodiment comprises,
the image acquisition unit 1 is arranged on the front side of the motor vehicle, and is used for acquiring and transmitting image information in front of the motor vehicle;
the processing unit 2 is arranged in the motor vehicle and is in signal connection with the image acquisition unit 1;
the processing unit 2 is comprised of a processing unit,
an insulating case 20;
the sucker seat 21 is detachably connected with the interior of the motor vehicle, and the top of the sucker seat 21 is connected with the bottom of the insulating shell 20;
a display screen 22 embedded in one end face of the insulating housing 20;
a speaker 23 fixed to one side surface of the insulating housing 20;
the raspberry pi 24 is fixed inside the insulating shell 20, an input interface of the raspberry pi 24 is in line connection with the image acquisition unit 1, and an output interface of the raspberry pi 24 is in line connection with the display screen 22 and the loudspeaker 23 respectively;
the raspberry pi 24 is a raspberry pi carrying a deep convolutional neural network algorithm.
The existing vehicle-mounted device for identifying and detecting the traffic sign is high in cost, a plurality of structural parts need to be additionally arranged and modified on a motor vehicle, and a user needs to pay a large cost for transferring the traffic identification and detection device after replacing the motor vehicle.
As shown in fig. 1, in this embodiment, image information in front of a motor vehicle acquired by an image acquisition unit 1 is transmitted to a processing unit 2 for identification processing, a display screen 22 for displaying images and a speaker 23 for broadcasting road traffic information on the processing unit 2 are integrally installed on an insulating housing 20, the insulating housing 20 is detachably connected with the inside of the motor vehicle through a suction cup seat 21, the suction cup seat 21 can be attached to an instrument desk side beside a driver, and the assembly and disassembly are both convenient; the image information acquired by the image acquisition unit 1 is processed by the raspberry 24 and then presented to the driver by the loudspeaker 23 and the display screen 22 in an audible and visual manner.
Further, the raspberry pi 24 of the present embodiment is a raspberry pi carrying a deep convolutional neural network algorithm, which has great advantages in the field of image recognition and object recognition due to its better feature learning ability and better characteristics, the convolutional neural network adopted by the raspberry pi 24 of the present embodiment is Mask R-CNN, and the cascade Box Head is adopted to perform multi-stage optimization on the input candidate frame; the network mainly comprises three parts: the prediction method comprises the steps that a main network, a regional generation network and a prediction network are selected, the main network is an improved FPN structure of a ResNet series network, a residual error unit is introduced into the ResNet network, the structure adopts a structure that a 3x3 convolution layer is added between two 1x1 convolution layers, the main function of the regional generation network is to generate a candidate frame for prediction of a subsequent Head network, the prediction network mainly comprises three cascaded Box heads and a Mask Head, and the candidate frame generated by the prediction network is optimized step by step through the three Box heads to finally generate a high-quality prediction frame; and then, through example segmentation prediction of a Mask Head generated target, as shown in fig. 9, the raspberry pi 24 carrying the deep convolutional neural network algorithm can completely meet the image level detection task, so that the recognition speed of the device is improved, and the device has the advantages of high detection precision, wide application scene and the like.
The raspberry pi 24 of this embodiment reads the image information collected by the image collecting unit 1 through the CSI interface, analyzes and processes the information in the trained deep convolutional neural network, determines whether or not there is a traffic sign in the image, and what the traffic sign is, and the fed back graphic information is presented to the driver through the display screen 22 and is broadcasted through the speaker 23 by voice.
The traffic sign recognition and detection device based on the deep convolutional neural network of the embodiment, the structure is simple, the installation and the disassembly are all convenient, the private car is replaced with the traffic sign at the current with frequent iteration, the transfer cost of the device is very low, the device can show the detection and recognition result of the traffic sign to a driver more comprehensively in two modes, namely hearing and vision, the device adopts the raspberry with the deep convolutional neural network algorithm to send 24 to detect and recognize images, the recognition speed is high, the detection precision is high, and the application scene is wide.
Example 2
The traffic sign recognition and detection device based on the deep convolutional neural network of the embodiment is further improved on the basis of the embodiment 1, and the image acquisition unit 1 includes:
a base 10 fixed to a front side of the motor vehicle;
the camera 11 is arranged above the base 10, the camera 11 collects image information in front of the motor vehicle, and a polaroid 110 is arranged on a lens of the camera 11.
In this embodiment, the camera 11 is adopted to collect image information, more specifically, the camera 11 of 1080P is adopted, and interference factors such as reflection of light are reduced by installing the polarizer 110 on the lens, and more clear and high-quality image information is collected to be provided for the raspberry group 24 to perform identification analysis.
Example 3
The traffic sign recognition and detection device based on the deep convolutional neural network of the embodiment is further improved on the basis of the embodiments 1 and 2, and the image acquisition unit 1 further includes:
a rotary support 12 fixed to the top surface of the base 10;
and one end of the adjustable hose 13 is fixedly connected with the top of the rotating support 12, and the other end of the adjustable hose is fixedly connected with the camera 11.
The bottom and the base 10 top surface fixed connection of rotating support 12, the top can be rotatory around the bottom, and adjustable hose 13 is hollow structure, and the shape position of adjustable hose 13 self can be adjusted in a flexible way, and this embodiment is through fixing camera 11 on adjustable hose 13, and the angular position of adjustment camera 11 that can be nimble comes better acquisition image information so that adapt to different motorcycle types, driving environment.
Further, a USB data line is led out from the bottom of the camera 11, passes through the adjustable hose 13, and is finally connected with an input interface line of the raspberry pi 24 in the insulating housing 20.
Example 4
The traffic sign recognition and detection device based on the deep convolutional neural network is further improved on the basis of the embodiments 1 to 3, and a relief port 200 is formed in the other end face, opposite to the display screen 22, of the insulating shell 20;
also included is an adjustable support mechanism 25, which includes:
the adjusting piece 250 is arranged in the insulating housing 20, the adjusting piece 250 comprises a transverse plate 2500 and two side plates 2501, the transverse plate 2500 is horizontally arranged, the two side plates 2501 are parallel to each other and are respectively fixed at two ends of the transverse plate 2500, one side of each side plate 2501 is fixed with the end part of the transverse plate 2500, and a tooth-shaped protrusion 2502 is formed at the bottom of the other side of each side plate 2501;
the supporting member 251 is arranged below the adjusting member 250 and corresponds to the position of the relief opening 200, the supporting member 251 comprises two supporting rods 2510 which are respectively and correspondingly arranged below the two side plates 2501, a connecting plate 2511 is vertically arranged on one side of the top surface of the supporting rod 2510, the connecting plate 2511 is rotatably connected with the side plates 2501, and a tooth-shaped recess 2512 which is matched and connected with the tooth-shaped protrusion 2502 is formed on the other side of the top surface of the supporting rod 2510.
As shown in fig. 2 and 3, in the present embodiment, the bottom of the insulating housing 20 is hinged to the top of the suction cup seat 21, so as to facilitate adjusting the angle of the insulating housing 20, so as to adjust the orientation of the display screen 22, and make it convenient for the user to observe at a more proper angle, the present embodiment is provided with an adjustable supporting mechanism 25, as shown in fig. 4, after the display screen 22 is adjusted to a proper orientation, the insulating housing 20 is supported at the other end of the insulating housing 20.
In this embodiment, the avoiding opening 200 is formed at the lower portion of the end surface of the insulating housing 20 opposite to the display screen 22, the adjusting member 250 is connected to the inner wall of the end surface of the insulating housing 20 where the avoiding opening 200 is located, and the supporting member 251 can extend out of the avoiding opening 200 at a position corresponding to the avoiding opening 200.
As shown in fig. 5 and 6, the horizontal plate 2500 of the adjusting member 250 is fixed on one side of the side plate 2501, faces the inside of the insulating housing 20, and the other side faces the relief opening 200, and the tooth-shaped protrusion 2502 is formed at the bottom of the other side, and a limiting plate 203 is formed on the inner wall of the end face of the insulating housing 20 where the relief opening 200 is located, and the adjusting member 250 is placed on the limiting plate 203.
As shown in fig. 7 and 8, the supporting member 251 is provided with a connecting plate 2511 on the top surface of the supporting member 251 located between two transverse plates 2500 of the adjusting member 250, pin holes are respectively formed in the connecting plate 2511 and the transverse plates 2500, a connecting pin passes through the pin holes to rotatably connect the transverse plates 2500 with the corresponding connecting plates 2511, and two supporting rods 2510 of the supporting member 251 are connected through a reinforcing rod 2513 to increase the strength of the supporting rods 2510.
The tooth-shaped protrusion 2502 of the adjusting member 250 is engaged with the tooth-shaped recess 2512 of the supporting member 251, when the insulating housing 20 is vertically disposed, the supporting member 251 is in a vertical position, as shown in fig. 4, after the angle of the insulating housing 20 is adjusted, the supporting member 251 rotates around the transverse plate 2500, the tooth-shaped recess 2512 and the tooth-shaped protrusion 2502 rotate in a dislocation manner, and after the supporting member 251 rotates to a proper position, the supporting member 251 can be kept at a stable position by the engagement between the tooth-shaped recess 2512 and the tooth-shaped protrusion 2502, so that a supporting effect is achieved.
Further, at least one of the tooth-shaped protrusion 2502 and the tooth-shaped recess 2512 is an elastic structure, and the elastic structure facilitates the dislocation rotation of the tooth-shaped recess 2512 and the tooth-shaped protrusion 2502 when the supporting member 251 rotates and adjusts.
Example 5
The traffic sign recognition and detection device based on the deep convolutional neural network is further improved on the basis of embodiments 1 to 4, a guide rail 202 is arranged on the inner wall above the escape opening 200 of the insulating shell 20, an adjusting piece 250 is connected in the guide rail 202 in a sliding manner, and a lifting opening 201 is formed in the top surface of the insulating shell 20 right above the guide rail 202;
the lifting handle 252 is fixed on the top of the two connecting plates 2511, and the lifting handle 252 is positioned right below the lifting handle opening 201.
As shown in fig. 3 and 4, the adjusting member 250 is slidably disposed in the guide rail 202, the bottom of the guide rail 202 is a limiting plate 203 for limiting the lowest position of the adjusting member 250, when the angle adjustment support is performed, the adjusting member 250 is located at the lowest position, when it is necessary to move the position of the device, the supporting member 251 can be restored to the initial position, and then the handle 252 is extended from the handle opening 201, thereby increasing the portability of the device.
The working principle and the working process of the device are as follows:
fixing a base 10 of an image acquisition unit 1 in front of a motor vehicle driving seat, adjusting a 1080P camera 11 to a proper position angle, acquiring RGB image information in front of the motor vehicle, fixing an insulating shell 20 of a processing unit 2 in the motor vehicle through a sucker seat 21, reading back the information acquired by the camera 11 through a CSI interface by a raspberry group 24, analyzing and processing the information in a trained deep convolutional neural network, judging whether a traffic sign exists in the image or not and what traffic sign exists in the image, displaying the fed-back graphic information to a driver through a display screen 22, and playing the graphic information in voice through a loudspeaker 23; when the angle of the display screen 22 needs to be adjusted, the supporting member 251 is taken out of the abdicating opening 200 and rotated to a proper position around the adjusting member 250, and the insulating housing 20 is stably supported under the mutual meshing between the tooth-shaped recess 2512 and the tooth-shaped protrusion 2502; when the device needs to be disassembled or transferred, the supporting member 251 is retracted into the position-allowing opening 200, and the supporting member 251 is pushed upwards, so that the handle 252 extends out of the handle opening 201 at the top of the insulating shell 20, the whole processing unit 2 is in a small suitcase type structure, the sucker seat 21 is disconnected from the interior of the motor vehicle, and the processing unit 2 can be moved.
The examples of the utility model are only right the utility model discloses a preferred embodiment describes, and not right the utility model discloses design and scope are injectd, do not deviate from the utility model discloses under the prerequisite of design idea, the field engineering technical personnel are right the utility model discloses a various deformation and improvement that technical scheme made all should fall into the protection scope of the utility model.
Claims (6)
1. A traffic sign identification and detection device based on a deep convolutional neural network is characterized by comprising,
the image acquisition unit is arranged on the front side of the motor vehicle, and is used for acquiring and transmitting image information in front of the motor vehicle;
the processing unit is arranged in the motor vehicle and is in signal connection with the image acquisition unit;
the processing unit comprises a processing unit and a control unit,
an insulating housing;
the sucker seat is detachably connected with the interior of the motor vehicle, and the top of the sucker seat is connected with the bottom of the insulating shell;
the display screen is embedded in one end face of the insulating shell;
a speaker fixed to one side surface of the insulating housing;
the raspberry pie is fixed inside the insulating shell, an input interface of the raspberry pie is connected with the image acquisition unit through a wire, and an output interface of the raspberry pie is respectively connected with the display screen and the loudspeaker through a wire;
the raspberry pie is a raspberry pie loaded with a deep convolutional neural network algorithm.
2. The traffic sign recognition and detection device based on the deep convolutional neural network as claimed in claim 1, wherein the image acquisition unit comprises:
a base fixed to a front side of the motor vehicle;
the camera is arranged above the base and used for collecting image information in front of the motor vehicle, and a polaroid is arranged on the camera lens.
3. The traffic sign recognition and detection device based on the deep convolutional neural network as claimed in claim 2, wherein the image acquisition unit further comprises:
the rotating support is fixed on the top surface of the base;
and one end of the adjustable hose is fixedly connected with the top of the rotary support, and the other end of the adjustable hose is fixedly connected with the camera.
4. The traffic sign recognition and detection device based on the deep convolutional neural network as claimed in claim 1, wherein a relief port is formed on the other end surface of the insulating shell opposite to the display screen;
still include adjustable supporting mechanism, it includes:
the adjusting piece is arranged in the insulating shell and comprises a transverse plate and two side plates, the transverse plate is horizontally arranged, the two side plates are parallel to each other and are respectively fixed at two ends of the transverse plate, one side of each side plate is fixed with the end part of the transverse plate, and a tooth-shaped bulge is formed at the bottom of the other side of each side plate;
the supporting piece is arranged below the adjusting piece and corresponds to the position of the abdicating opening, the supporting piece comprises two supporting rods which are respectively arranged below the two side plates correspondingly, one side of the top surface of each supporting rod is vertically provided with a connecting plate, the connecting plate is rotatably connected with the side plates, and the other side of the top surface of each supporting rod is provided with a tooth-shaped recess which is matched and connected with the tooth-shaped protrusion.
5. The traffic sign recognition and detection device based on the deep convolutional neural network as claimed in claim 4, wherein a guide rail is arranged on the inner wall above the abdication port of the insulating shell, the adjusting piece is connected in the guide rail in a sliding manner, and a handle port is formed on the top surface of the insulating shell right above the guide rail;
the lifting handle is fixed on the tops of the two connecting plates and is positioned right below the lifting handle opening.
6. The traffic sign recognition and detection device based on the deep convolutional neural network as claimed in claim 4, wherein: at least one of the tooth-shaped protrusion and the tooth-shaped recess is of an elastic structure.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202120759449.1U CN214751919U (en) | 2021-04-14 | 2021-04-14 | Traffic sign recognition and detection device based on deep convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202120759449.1U CN214751919U (en) | 2021-04-14 | 2021-04-14 | Traffic sign recognition and detection device based on deep convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN214751919U true CN214751919U (en) | 2021-11-16 |
Family
ID=78602931
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202120759449.1U Expired - Fee Related CN214751919U (en) | 2021-04-14 | 2021-04-14 | Traffic sign recognition and detection device based on deep convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN214751919U (en) |
-
2021
- 2021-04-14 CN CN202120759449.1U patent/CN214751919U/en not_active Expired - Fee Related
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105182535B (en) | The method that automobile maintenance is carried out using intelligent glasses | |
| WO2021179679A1 (en) | Automobile dashboard testing system and testing method based on machine vision | |
| CN105030205B (en) | Handheld optical high-resolution high-power intelligent skin testing analyzer and system and method thereof | |
| CN104717465B (en) | Wide-angle video monitoring equipment and video monitoring method for multiple IP cameras | |
| US8189181B2 (en) | Method and system of evaluating a stereo image displaying panel | |
| CN107767377B (en) | Liquid crystal display defect and dust distinguishing method and detection device based on binocular vision system | |
| CN105243386A (en) | Face living judgment method and system | |
| CN208207914U (en) | PCB surface defect detecting device based on fast geometric alignment | |
| CN110940319A (en) | Height limit detection method and device based on machine vision recognition | |
| CN103581649A (en) | Naked eye three-dimensional image display device and method | |
| CN214751919U (en) | Traffic sign recognition and detection device based on deep convolutional neural network | |
| CN113752275A (en) | AI device capable of automatically identifying defects of substation equipment | |
| CN200994176Y (en) | Front road scene collecting platform | |
| CN206563561U (en) | Mobile phone housing hole automatic checkout equipment | |
| CN117761060A (en) | Visual detection system and detection method thereof | |
| CN108414961B (en) | Detection device and detection method for electric wire of ammeter | |
| JP7067852B1 (en) | Calculation method of road surface damage position | |
| CN115471482A (en) | Small-caliber container inner wall defect detection method based on computer vision | |
| CN109374637A (en) | A kind of automobile panel Surface Quality Inspection System and method | |
| CN110688876A (en) | Lane line detection method and device based on vision | |
| CN109003273A (en) | A kind of car light light guide consistency detecting method | |
| CN115901750A (en) | Glass detection method | |
| CN112033973A (en) | Big data intelligent analysis-based second-hand automobile performance detection device | |
| CN204492801U (en) | A kind of probe of boring imaging system of adjustable focus | |
| CN110728281A (en) | License plate segmentation and recognition method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20211116 |