CN208903309U - Integrate the object detection and range unit of deep learning Yu binocular ranging - Google Patents
Integrate the object detection and range unit of deep learning Yu binocular ranging Download PDFInfo
- Publication number
- CN208903309U CN208903309U CN201821397790.1U CN201821397790U CN208903309U CN 208903309 U CN208903309 U CN 208903309U CN 201821397790 U CN201821397790 U CN 201821397790U CN 208903309 U CN208903309 U CN 208903309U
- Authority
- CN
- China
- Prior art keywords
- output end
- image
- camera
- unit
- processing unit
- 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 15
- 238000013135 deep learning Methods 0.000 title claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 8
- 238000005259 measurement Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
Integrate the object detection and range unit of deep learning Yu binocular ranging, it is related to computer vision process field, more particularly to the object detection and range unit for integrating deep learning Yu binocular ranging, it includes power supply unit, left camera, right camera, image input port, central processing unit, image processing unit, computing unit, memory, output end of image mouth, display screen, power supply unit is electrically connected with left camera and right camera respectively, the output end of left camera and right camera is connect with the input terminal of image input port, the output end of image input port and the input terminal of central processing unit connect, the output end of memory is connect with the input terminal of output end of image mouth, the output end of output end of image mouth and the input terminal of display screen connect.After adopting the above technical scheme, the utility model has the beneficial effect that it can not only reduce cost, camera can also be made to become more intelligent, functional diversities, the utilization rate for improving camera.
Description
Technical field
The utility model relates to computer vision process fields, and in particular to integrate deep learning and binocular ranging
Object detection and range unit.
Background technique
Currently, image procossing is a very active field, and ranging technology is also essential in practical applications.Make
Additional cost is not only brought with additional range unit, but also may need to occupy certain hardware resources, such as in monolithic
Additional I/O mouth, interrupt resources etc. can be occupied by carrying out ranging using ultrasonic equipment or Laser Distance Measuring Equipment on machine, arm.And
It transfers to image processing section to realize ranging task, then can be effectively solved the above problem.Deep learning is in object detection
The important role of performer combines if function can be will test with ranging, and object detection, distance measurement function are transferred to a binocular
Camera is completed, and can not only reduce the cost of required hardware in engineering, camera can also be made to become more intelligent, vdiverse in function
Change, improve the utilization rate of camera.
Binocular ranging scheme is applied to different scenes mostly by existing implementation, only the single mesh for realizing ranging
, or the matching algorithm of binocular ranging is improved, to improve the precision of ranging.Existing binocular ranging technology is only real
The distance measurement function of certain existing point, can not detect the object in image.
Utility model content
The purpose of this utility model is that it is in view of the drawbacks of the prior art and insufficient, collection deep learning and binocular ranging are provided
In the object detection and range unit of one, it can will test function and combine with ranging, and object detection, distance measurement function are transferred to
One binocular camera is completed, and the cost of required hardware in engineering can not only be reduced, can also make camera become it is more intelligent,
Functional diversities, the utilization rate for improving camera.
To achieve the above object, the utility model is using following technical scheme: it includes power supply unit 1, left camera
2, right camera 3, image input port 4, central processing unit 5, image processing unit 51, computing unit 52, memory 6, image
Output port 7, display screen 8, power supply unit 1 are electrically connected with left camera 2 and right camera 3 respectively, and left camera 2 and the right side are taken the photograph
As first 3 output end is connect with the input terminal of image input port 4, the output end and central processing unit 5 of image input port 4
Input terminal connection, central processing unit 5 includes image processing unit 51 and computing unit 52, image processing unit 51 and is calculated single
Member 52 is arranged in parallel, and the output end of image processing unit 51 and computing unit 52 is connect with the input terminal of memory 6, memory
6 output end is connect with the input terminal of output end of image mouth 7, and the output end of output end of image mouth 7 and the input terminal of display screen 8 connect
It connects.
The left camera 2 and right camera 3 is linearly distributed in same level.
The distance between described left camera 2 and right camera 3 are adjustable.Adopting the structure can guarantee that two are taken the photograph
As head acquires original image with different view.
After adopting the above technical scheme, the utility model has the beneficial effect that it can will test function and combine with ranging, it will
Object detection, distance measurement function transfer to a binocular camera to complete, and can not only reduce the cost of required hardware in engineering, also can
Camera is set to become more intelligent, functional diversities, the utilization rate for improving camera.
Detailed description of the invention
In order to illustrate the embodiment of the utility model or the technical proposal in the existing technology more clearly, below will be to embodiment
Or attached drawing needed to be used in the description of the prior art is briefly described, it should be apparent that, the accompanying drawings in the following description is only
It is some embodiments of the utility model, for those of ordinary skill in the art, before not making the creative labor property
It puts, is also possible to obtain other drawings based on these drawings.
Fig. 1 is the structural schematic block diagram of the utility model.
Description of symbols: power supply unit 1, left camera 2, right camera 3, image input port 4, central processing unit 5,
Image processing unit 51, computing unit 52, memory 6, output end of image mouth 7, display screen 8.
Specific embodiment
Referring to shown in Fig. 1, present embodiment the technical solution adopted is that: it include power supply unit 1, left camera 2,
Right camera 3, image input port 4, central processing unit 5, image processing unit 51, computing unit 52, memory 6, image are defeated
Exit port 7, display screen 8, power supply unit 1 are electrically connected with left camera 2 and right camera 3 respectively, left camera 2 and right camera shooting
First 3 output end is connect with the input terminal of image input port 4, output end and the central processing unit 5 of image input port 4
Input terminal connection, central processing unit 5 include image processing unit 51 and computing unit 52, image processing unit 51 and computing unit
52 are arranged in parallel, and the output end of image processing unit 51 and computing unit 52 is connect with the input terminal of memory 6, memory 6
Output end connect with the input terminal of output end of image mouth 7, the input terminal of the output end of output end of image mouth 7 and display screen 8 connects
It connects.
The left camera 2 and right camera 3 is linearly distributed in same level.
The distance between described left camera 2 and right camera 3 are adjustable.Adopting the structure can guarantee that two are taken the photograph
As head acquires original image with different view.
Working principle of the utility model is: during realizing object detection, ranging using binocular camera, it is sharp first
With binocular camera acquisition reference object image (acquire multipair original image from each visual angle of calibration object and different distance,
Image is as much as possible as more as possible to obtain enough scenes comprising information, the quantity of image such as different distance, angles
Information).Binocular camera is demarcated using acquired image, gets the parameter of binocular camera, and utilize these
Parameter corrects the image acquired in real time in subsequent use process.Then using on image after deep learning detection correction
Object, and ranging to detected target is realized according to binocular range measurement principle.
The above is merely intended for describing the technical solutions of the present application, but not for limiting the present application, those of ordinary skill in the art couple
The other modifications or equivalent replacement that the technical solution of the utility model is made, without departing from technical solutions of the utility model
Spirit and scope should all cover in the scope of the claims of the utility model.
Claims (3)
1. integrating the object detection and range unit of deep learning Yu binocular ranging, it is characterised in that: it includes that power supply is single
First (1), left camera (2), right camera (3), image input port (4), central processing unit (5), image processing unit (51),
Computing unit (52), memory (6), output end of image mouth (7), display screen (8), power supply unit (1) respectively with left camera (2)
It is electrically connected with right camera (3), the output end of left camera (2) and right camera (3) is defeated with image input port (4)
Enter end connection, the output end of image input port (4) is connect with the input terminal of central processing unit (5), and central processing unit (5) includes
Image processing unit (51) and computing unit (52), image processing unit (51) and computing unit (52) are arranged in parallel, at image
Reason unit (51) and the output end of computing unit (52) are connect with the input terminal of memory (6), the output end of memory (6) and
The input terminal of output end of image mouth (7) connects, and the output end of output end of image mouth (7) is connect with the input terminal of display screen (8).
2. the object detection and range unit according to claim 1 for integrating deep learning Yu binocular ranging, special
Sign is: the left camera (2) and right camera (3) is linearly distributed in same level.
3. the object detection and range unit according to claim 1 for integrating deep learning Yu binocular ranging, special
Sign is: the distance between described left camera (2) and right camera (3) are adjustable.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201821397790.1U CN208903309U (en) | 2018-08-29 | 2018-08-29 | Integrate the object detection and range unit of deep learning Yu binocular ranging |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201821397790.1U CN208903309U (en) | 2018-08-29 | 2018-08-29 | Integrate the object detection and range unit of deep learning Yu binocular ranging |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN208903309U true CN208903309U (en) | 2019-05-24 |
Family
ID=66570587
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201821397790.1U Expired - Fee Related CN208903309U (en) | 2018-08-29 | 2018-08-29 | Integrate the object detection and range unit of deep learning Yu binocular ranging |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN208903309U (en) |
-
2018
- 2018-08-29 CN CN201821397790.1U patent/CN208903309U/en not_active Expired - Fee Related
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104215217B (en) | A kind of pseudo- image distance monocular distance-finding method applied to power-line patrolling unmanned plane | |
| CN109801314B (en) | Binocular dynamic vision sensor stereo matching method based on deep learning | |
| CN104089697A (en) | Real-time online visual vibration measurement method based on thread pool concurrent technology | |
| CN106537087A (en) | Real-time video extensometer | |
| CN107766812A (en) | A kind of real-time face detection identifying system based on MiZ702N | |
| CN107144703A (en) | Embedded image acquisition and processing system and method based on particle image velocimetry | |
| CN112256627A (en) | High-definition high-frame-frequency real-time image processing platform and method for photoelectric tracking system | |
| CN105380592A (en) | Wearable equipment for pupil distance detection and implementation method thereof | |
| CN103322980A (en) | Fruit attitude detection method based on stereoscopic vision | |
| CN111860042A (en) | Method and apparatus for meter reading | |
| CN208903309U (en) | Integrate the object detection and range unit of deep learning Yu binocular ranging | |
| CN109063717A (en) | A kind of acquisition instrument center point method | |
| CN106152946B (en) | A kind of method and terminal measuring object length | |
| CN109121194B (en) | Method and apparatus for state transition of electronic device | |
| CN118351143A (en) | A low-power multi-point image displacement online tracking measurement system, method and medium | |
| CN207423308U (en) | A kind of digital water gauge mark etc. divides measuring system | |
| CN105447007A (en) | Electronic device and data processing method | |
| CN109246417A (en) | A kind of machine vision analysis system and method based on bore hole stereoscopic display | |
| CN204206306U (en) | A kind of intelligent electric meter element image capture device | |
| CN110210322A (en) | A method of recognition of face is carried out by 3D principle | |
| CN107796319A (en) | A kind of wire measuring device and its measuring method | |
| CN101777182A (en) | Video positioning method of coordinate cycling approximation type orthogonal camera system and system thereof | |
| CN105651170B (en) | The detection system and its detection method of flow position | |
| CN209170524U (en) | A kind of 3D structure optical mode group | |
| CN101598846B (en) | Computer terminal display system and display method for zoom lens target ranging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190524 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |