CN219571464U - Visual acquisition device based on deep neural network training - Google Patents
Visual acquisition device based on deep neural network training Download PDFInfo
- Publication number
- CN219571464U CN219571464U CN202320938421.3U CN202320938421U CN219571464U CN 219571464 U CN219571464 U CN 219571464U CN 202320938421 U CN202320938421 U CN 202320938421U CN 219571464 U CN219571464 U CN 219571464U
- Authority
- CN
- China
- Prior art keywords
- shaped plate
- camera
- neural network
- acquisition device
- device based
- 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.)
- Active
Links
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The utility model relates to the technical field of photogrammetry, in particular to a vision acquisition device based on deep neural network training, which comprises a camera and an adjusting component, wherein the adjusting component comprises a rotating mechanism, a bracket, a U-shaped plate, a clamping mechanism, a sliding block, a screw rod, a first motor and a connecting frame; the camera is arranged on the inner side of the U-shaped plate by the clamping mechanism, the lens of the camera is positioned on one side far away from the connecting frame, the U-shaped plate can rotate on the bracket, the rotating mechanism drives the bracket to rotate, the angle of the camera in the horizontal direction is adjusted, the first motor drives the screw to rotate, the screw drives the sliding block to slide, and the sliding block drives the U-shaped plate to rotate through the connecting frame; when the sliding block slides upwards, the connecting frame jacks up the left side of the U-shaped plate, so that the lens of the camera moves downwards, when the sliding block slides downwards, the connecting frame pulls the left side of the U-shaped plate downwards, so that the lens of the camera moves upwards, the pitching angle of the camera can be adjusted, and the blind area of the visual field is reduced.
Description
Technical Field
The utility model relates to the technical field of photogrammetry, in particular to a vision acquisition device based on deep neural network training.
Background
The existing visual acquisition device based on deep neural network training is used for acquiring images through a camera, but the camera is fixed in position and inconvenient to rotate in the horizontal direction for adjusting the angle.
The prior art provides a vision collection system based on degree of depth neural network training, has rotated on the base and has installed a supporting seat, has installed the camera on the supporting seat, rotates the supporting seat and can drive the camera and rotate in the horizontal direction, adjusts collection angle.
But adopt above-mentioned mode, the camera is fixed mounting on the supporting seat, can't adjust every single move angle on the supporting seat, leads to appearing a large amount of visual field blind areas.
Disclosure of Invention
The utility model aims to provide a visual acquisition device based on deep neural network training, which can adjust the pitching angle of a camera and reduce the blind area of a visual field.
In order to achieve the above purpose, the utility model provides a vision acquisition device based on deep neural network training, which comprises a camera and an adjusting component;
the adjusting component comprises a rotating mechanism, a bracket, a U-shaped plate, a clamping mechanism, a sliding block, a screw rod, a first motor and a connecting frame; the rotating mechanism is arranged below the camera, the bracket is arranged above the rotating mechanism, the U-shaped plate is rotationally connected with the bracket and is positioned above the bracket, the camera is positioned on the inner side of the U-shaped plate, the clamping mechanism is arranged on the U-shaped plate, the sliding block is in sliding connection with the bracket and is penetrated by the bracket, the screw rod with the support rotates to be connected, and with slider threaded connection, and passes the slider, first motor with support fixed connection, and be located the support is inboard, the output shaft of first motor with screw rod fixed connection, the link with the slider rotates to be connected, and with U-shaped board rotates to be connected, and is located U-shaped board below.
The rotating mechanism comprises a base, a turntable and a second motor, wherein the base is located below the support, the turntable is rotationally connected with the base and fixedly connected with the support, the turntable is located between the base and the support, the second motor is fixedly connected with the base and located on the inner side of the base, and an output shaft of the second motor is fixedly connected with the turntable.
The clamping mechanism comprises a clamping plate, a first stud and a knob, wherein the clamping plate is in sliding connection with the U-shaped plate and is positioned on the inner side of the U-shaped plate, the first stud is in rotational connection with the clamping plate and is in threaded connection with the U-shaped plate and penetrates through the U-shaped plate, and the knob is fixedly connected with the first stud and is positioned on one side of the first stud away from the clamping plate.
The clamping mechanism further comprises a buffer cushion, wherein the buffer cushion is fixedly connected with the clamping plate and is positioned on one side, away from the first stud, of the clamping plate.
The visual acquisition device based on the deep neural network training further comprises a limiting component, and the limiting component is arranged on the U-shaped plate.
The limiting assembly comprises a first baffle, a mounting block, a sliding rod, a second baffle and a second stud, wherein the first baffle is fixedly connected with the U-shaped plate and is positioned on one side of the U-shaped plate, the mounting block is fixedly connected with the U-shaped plate and is positioned on one side of the U-shaped plate, the sliding rod is in sliding connection with the mounting block and penetrates through the mounting block, the second baffle is fixedly connected with the sliding rod and is positioned on one side of the sliding rod, and the second stud is in threaded connection with the mounting block and is in rotary connection with the second baffle and penetrates through the second baffle.
The limiting assembly further comprises a limiting block, wherein the limiting block is fixedly connected with the sliding rod and located on one side, away from the second baffle, of the sliding rod.
According to the vision acquisition device based on deep neural network training, the camera is clamped and mounted on the inner side of the U-shaped plate by the clamping mechanism, the lens of the camera is located on one side far away from the connecting frame, the U-shaped plate can rotate on the support, the rotating mechanism is used for driving the support to rotate and finally driving the camera to rotate in the horizontal direction, the angle of the camera in the horizontal direction is adjusted, the first motor is used for driving the screw to rotate, the screw rotates to drive the sliding block to slide, and the sliding block drives the U-shaped plate to rotate through the connecting frame; the method comprises the following steps: when the sliding block slides upwards, the connecting frame jacks up the left side of the U-shaped plate, so that the lens of the camera moves downwards, when the sliding block slides downwards, the connecting frame pulls the left side of the U-shaped plate downwards, so that the lens of the camera moves upwards, the pitching angle of the camera can be adjusted, and the blind area of the visual field is reduced.
Drawings
In order to more clearly illustrate the embodiments of the present utility model or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
Fig. 1 is a schematic overall structure of a first embodiment of the present utility model.
Fig. 2 is a top view of the entirety of a first embodiment of the present utility model.
Fig. 3 is an overall front view of a second embodiment of the present utility model.
101-camera, 102-adjustment assembly, 103-rotation mechanism, 104-bracket, 105-U-shaped plate, 106-clamping mechanism, 107-slider, 108-screw, 109-first motor, 110-link, 111-base, 112-carousel, 113-second motor, 114-splint, 115-first stud, 116-knob, 117-cushion, 201-spacing assembly, 202-first baffle, 203-mounting block, 204-slide bar, 205-second baffle, 206-second stud, 207-stopper.
Detailed Description
The first embodiment of the utility model is as follows:
referring to fig. 1-2, fig. 1 is a schematic overall structure of a first embodiment of the present utility model, and fig. 2 is a top view of the first embodiment of the present utility model. The utility model provides a visual acquisition device based on deep neural network training, which comprises: comprising a camera 101 and an adjustment assembly 102; the adjusting assembly 102 comprises a rotating mechanism 103, a bracket 104, a U-shaped plate 105, a clamping mechanism 106, a sliding block 107, a screw 108, a first motor 109 and a connecting frame 110; the rotating mechanism 103 comprises a base 111, a turntable 112 and a second motor 113; the clamping mechanism 106 includes a clamp plate 114, a first stud 115, a knob 116, and a buffer pad 117.
For the specific implementation mode, the neural network is an algorithm mathematical model for simulating the behavior characteristics of the animal neural network and carrying out distributed parallel information processing, and the network is used for achieving the purpose of processing information by adjusting the interconnection relation among a large number of internal nodes according to the complexity of the system. The depth neural network is one of the most representative models in the artificial intelligence field, the calculation process is simple and easy to understand, the fitting capability to data is strong, the extremely excellent performance is shown on a plurality of different tasks, the machine vision is the application based on the training of the depth neural network, the machine vision is to replace a human eye to measure and judge, and a typical machine vision application system comprises an image capturing module, a light source system, an image digitizing module, a digital image processing module, an intelligent judging and deciding module and a mechanical control executing module. In machine vision applications, cameras are a common type of image capture module. The camera 101 is a vision acquisition used in machine vision.
The camera 101 is located inside the U-shaped plate 105, the clamping mechanism 106 is disposed on the U-shaped plate 105, the sliding block 107 is slidably connected with the bracket 104 and is penetrated by the bracket 104, the screw 108 is rotatably connected with the bracket 104 and is in threaded connection with the sliding block 107 and penetrates through the sliding block 107, the first motor 109 is fixedly connected with the bracket 104 and is located inside the bracket 104, the output shaft of the first motor 109 is fixedly connected with the screw 108, the connecting frame 110 is rotatably connected with the sliding block 107 and is rotatably connected with the U-shaped plate 105 and is located below the U-shaped plate 105. The clamping mechanism 106 clamps and installs the camera 101 inside the U-shaped plate 105, the lens of the camera 101 is located at a side far away from the connecting frame 110, the U-shaped plate 105 can rotate on the bracket 104, the rotating mechanism 103 is used for driving the bracket 104 to rotate and finally driving the camera 101 to rotate in a horizontal direction, the angle of the camera 101 in the horizontal direction is adjusted, the first motor 109 is used for driving the screw 108 to rotate, the screw 108 rotates and drives the sliding block 107 to slide, and the sliding block 107 drives the U-shaped plate 105 to rotate through the connecting frame 110; the method comprises the following steps: when the sliding block 107 slides up, the connecting frame 110 pushes up the left side of the U-shaped plate 105, so that the lens of the camera 101 moves downwards, and when the sliding block 107 slides down, the connecting frame 110 pulls down the left side of the U-shaped plate 105, so that the lens of the camera 101 moves upwards, thereby adjusting the pitching angle of the camera 101 and reducing the blind area of the visual field.
Secondly, the base 111 is located below the support 104, the turntable 112 is rotatably connected with the base 111, and is fixedly connected with the support 104, and is located between the base 111 and the support 104, the second motor 113 is fixedly connected with the base 111, and is located inside the base 111, and an output shaft of the second motor 113 is fixedly connected with the turntable 112. The base 111 is rotatably provided with the turntable 112, the turntable 112 supports the bracket 104, the second motor 113 is used for driving the turntable 112 to rotate, thereby driving the bracket 104 to rotate, and finally driving the camera 101 to rotate in the horizontal direction, and adjusting the angle of the camera 101 in the horizontal direction.
Meanwhile, the clamping plate 114 is slidably connected with the U-shaped plate 105 and is located at the inner side of the U-shaped plate 105, the first stud 115 is rotatably connected with the clamping plate 114 and is in threaded connection with the U-shaped plate 105, and passes through the U-shaped plate 105, and the knob 116 is fixedly connected with the first stud 115 and is located at one side of the first stud 115 away from the clamping plate 114. When the camera 101 is installed, the camera 101 is placed in the U-shaped plate 105, one side of the camera 101 is close to the inner wall of the U-shaped plate 105, then the knob 116 is held, the first stud 115 is rotated to drive the clamping plate 114 to abut against the camera 101, and the clamping plate 114 is matched with the U-shaped plate 105 to clamp the camera 101 in the U-shaped plate 105.
In addition, the cushion 117 is fixedly connected to the clamping plate 114, and is located on a side of the clamping plate 114 away from the first stud 115. The housing of the camera 101 may be damaged when the clamping plate 114 abuts against the camera 101, so that the cushion pad 117 is provided, and the cushion pad 117 prevents the clamping plate 114 from rigidly colliding with the camera 101 when the clamping plate 114 abuts against the camera 101, resulting in damage to the camera 101.
According to the visual acquisition device based on deep neural network training, when the visual acquisition device is used, the camera 101 is placed into the U-shaped plate 105, one side of the visual acquisition device is close to the inner wall of the U-shaped plate 105, then the knob 116 is held, the first stud 115 is rotated, the clamping plate 114 is driven to tightly support the camera 101, the clamping plate 114 is matched with the U-shaped plate 105, the camera 101 is clamped into the U-shaped plate 105, when the horizontal angle of the camera 101 needs to be adjusted, the second motor 113 is started, the turntable 112 is driven to rotate, so that the bracket 104 is driven to rotate, finally the camera 101 is driven to rotate in the horizontal direction, the angle of the camera 101 is adjusted, when the camera 101 is required to be lifted up, the first motor 109 is started, the screw 108 is driven to rotate, the slider 107 is driven to slide down, the connecting frame 110 pulls the left side of the U-shaped plate 105 downwards, when the camera 101 needs to be lifted down, the first motor 109 is started, the camera 101 is driven to rotate in the opposite direction, and the screw 107 is driven to slide upwards.
The second embodiment of the utility model is as follows:
on the basis of the first embodiment, please refer to fig. 3, wherein fig. 3 is an overall front view of a second embodiment of the present utility model. The visual acquisition device based on deep neural network training provided by the utility model further comprises a limiting component 201; the limiting assembly 201 comprises a first baffle 202, a mounting block 203, a sliding rod 204, a second baffle 205, a second stud 206 and a limiting block 207.
For this embodiment, the spacing assembly 201 is disposed on the U-shaped plate 105. The camera 101 is held by the friction force between the cushion pad 117 and the U-shaped plate 105 and the camera 101, the camera 101 may slide down against the friction force in the process of adjusting the pitching angle, and the limiting component 201 is used for limiting the camera 101 to avoid sliding down.
The first baffle 202 is fixedly connected with the U-shaped plate 105, and is located at one side of the U-shaped plate 105, the mounting block 203 is fixedly connected with the U-shaped plate 105, and is located at one side of the U-shaped plate 105, the sliding rod 204 is slidably connected with the mounting block 203, and penetrates through the mounting block 203, the second baffle 205 is fixedly connected with the sliding rod 204, and is located at one side of the sliding rod 204, and the second stud 206 is in threaded connection with the mounting block 203, is in rotational connection with the second baffle 205, and penetrates through the second baffle 205. The first baffle 202 is used for limiting the camera 101 from the left side, the second baffle 205 can be moved to the right side through the cooperation of the mounting block 203 and the sliding rod 204, the second stud 206 is used for driving the second baffle 205 to move, when the camera 101 is mounted, the camera 101 is placed into the U-shaped plate 105, the left side of the camera 101 is tightly attached to the first baffle 202, then the second stud 206 is screwed, and the second stud 206 is screwed into the mounting block 203 to drive the second baffle 205 to slide to the left side until the second baffle 205 abuts against the right side of the camera 101.
Second, the limiting block 207 is fixedly connected with the sliding rod 204, and is located at a side of the sliding rod 204 away from the second baffle 205. The limiting block 207 is used for preventing the sliding rod 204 from being separated from the mounting block 203 without any limitation after the second stud 206 is separated from the mounting block 203.
According to the visual acquisition device based on deep neural network training, the limiting block 207 limits the limiting position of the sliding rod 204, when the camera 101 is installed, the camera 101 is placed into the U-shaped plate 105, the left side of the camera 101 is tightly attached to the first baffle 202, then the second stud 206 is screwed, the second stud 206 is screwed into the installation block 203 to drive the second baffle 205 to slide left, so that the second baffle 205 abuts against the right side of the camera 101, and then the camera 101 is clamped by the clamping mechanism, so that the camera 101 is stably installed in the U-shaped plate 105.
The foregoing disclosure is only illustrative of one or more preferred embodiments of the present utility model, and it is not intended to limit the scope of the claims hereof, as persons of ordinary skill in the art will understand that all or part of the processes for practicing the embodiments described herein may be practiced with equivalent variations in the claims, which are within the scope of the utility model.
Claims (7)
1. A vision acquisition device based on deep neural network training, which comprises a camera and is characterized in that,
the device also comprises an adjusting component;
the adjusting component comprises a rotating mechanism, a bracket, a U-shaped plate, a clamping mechanism, a sliding block, a screw rod, a first motor and a connecting frame;
the rotating mechanism is arranged below the camera, the bracket is arranged above the rotating mechanism, the U-shaped plate is rotationally connected with the bracket and is positioned above the bracket, the camera is positioned on the inner side of the U-shaped plate, the clamping mechanism is arranged on the U-shaped plate, the sliding block is in sliding connection with the bracket and is penetrated by the bracket, the screw rod with the support rotates to be connected, and with slider threaded connection, and passes the slider, first motor with support fixed connection, and be located the support is inboard, the output shaft of first motor with screw rod fixed connection, the link with the slider rotates to be connected, and with U-shaped board rotates to be connected, and is located U-shaped board below.
2. A vision acquisition device based on deep neural network training as claimed in claim 1,
the rotating mechanism comprises a base, a turntable and a second motor, wherein the base is positioned below the support, the turntable is rotationally connected with the base, is fixedly connected with the support, is positioned between the base and the support, is fixedly connected with the base and is positioned on the inner side of the base, and an output shaft of the second motor is fixedly connected with the turntable.
3. A vision acquisition device based on deep neural network training as claimed in claim 2,
the clamping mechanism comprises a clamping plate, a first stud and a knob, wherein the clamping plate is in sliding connection with the U-shaped plate and is positioned on the inner side of the U-shaped plate, the first stud is in rotational connection with the clamping plate and is in threaded connection with the U-shaped plate and penetrates through the U-shaped plate, and the knob is fixedly connected with the first stud and is positioned on one side of the first stud away from the clamping plate.
4. A visual acquisition device based on deep neural network training according to claim 3,
the clamping mechanism further comprises a buffer cushion, wherein the buffer cushion is fixedly connected with the clamping plate and is positioned on one side, away from the first stud, of the clamping plate.
5. A vision acquisition device based on deep neural network training as claimed in claim 1,
the visual acquisition device based on deep neural network training further comprises a limiting component, and the limiting component is arranged on the U-shaped plate.
6. A vision acquisition device based on deep neural network training as claimed in claim 5,
the limiting assembly comprises a first baffle, a mounting block, a sliding rod, a second baffle and a second stud, wherein the first baffle is fixedly connected with the U-shaped plate and is positioned on one side of the U-shaped plate, the mounting block is fixedly connected with the U-shaped plate and is positioned on one side of the U-shaped plate, the sliding rod is in sliding connection with the mounting block and penetrates through the mounting block, the second baffle is fixedly connected with the sliding rod and is positioned on one side of the sliding rod, and the second stud is in threaded connection with the mounting block and is in rotary connection with the second baffle and penetrates through the second baffle.
7. A vision acquisition device based on deep neural network training as claimed in claim 6,
the limiting assembly further comprises a limiting block, wherein the limiting block is fixedly connected with the sliding rod and is located on one side, away from the second baffle, of the sliding rod.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202320938421.3U CN219571464U (en) | 2023-04-24 | 2023-04-24 | Visual acquisition device based on deep neural network training |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202320938421.3U CN219571464U (en) | 2023-04-24 | 2023-04-24 | Visual acquisition device based on deep neural network training |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN219571464U true CN219571464U (en) | 2023-08-22 |
Family
ID=87657009
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202320938421.3U Active CN219571464U (en) | 2023-04-24 | 2023-04-24 | Visual acquisition device based on deep neural network training |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN219571464U (en) |
-
2023
- 2023-04-24 CN CN202320938421.3U patent/CN219571464U/en active Active
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105654001A (en) | Test method for two-dimensional code scanning equipment performance based on different field depths | |
| CN111735825A (en) | Pole piece edge detection device and detection method | |
| CN219571464U (en) | Visual acquisition device based on deep neural network training | |
| CN211857543U (en) | Ticket checking device based on face recognition | |
| CN211406101U (en) | Face recognition camera capable of improving recognition accuracy | |
| CN210834697U (en) | Gear surface defect detection device | |
| CN109185636A (en) | The motro drivien photographic device of adjustable direction | |
| CN210662071U (en) | Image acquisition device for image processing | |
| CN208607416U (en) | A kind of point light source communication device for optical camera image quality evaluating | |
| CN113558555A (en) | Oral cavity information acquisition device | |
| CN113757503B (en) | Facial biological feature acquisition method and acquisition device | |
| CN110908408A (en) | High follow-up automatic light following system capable of achieving quick positioning | |
| CN117030714B (en) | Mobile phone middle frame appearance detection mechanism and detection method | |
| CN221038729U (en) | Notebook keyboard outward appearance flaw detection machine | |
| CN115684172A (en) | Automobile appearance detection system and using method thereof | |
| CN116025819B (en) | On-site data acquisition system based on image recognition | |
| CN211821487U (en) | Face recognition device with adjustable recognition range | |
| CN219320793U (en) | Visual inspection device for surface defects of rudder structural members | |
| CN214751919U (en) | Traffic sign recognition and detection device based on deep convolutional neural network | |
| CN110824577A (en) | Infrared target detection platform and detection method thereof | |
| CN214337992U (en) | But camera light-resistant system | |
| CN213934929U (en) | Image recognition module based on deep neural network | |
| CN218917220U (en) | Multi-angle photographing device | |
| CN213544431U (en) | Appearance inspection machine and distance-adjustable imaging device thereof | |
| CN210745344U (en) | Video content acquisition and identification device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |