CN115115768A - Object coordinate recognition system, method, device and medium based on stereoscopic vision - Google Patents
Object coordinate recognition system, method, device and medium based on stereoscopic vision Download PDFInfo
- Publication number
- CN115115768A CN115115768A CN202110291250.5A CN202110291250A CN115115768A CN 115115768 A CN115115768 A CN 115115768A CN 202110291250 A CN202110291250 A CN 202110291250A CN 115115768 A CN115115768 A CN 115115768A
- Authority
- CN
- China
- Prior art keywords
- target object
- binocular camera
- stereoscopic vision
- dimensional
- camera
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30244—Camera pose
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention provides a stereoscopic vision-based object coordinate identification method, which is characterized by comprising the following steps of: step S1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system; step S2: respectively acquiring front and top images by using a multi-path camera and identifying a target object based on a convolutional neural network; step S3: after the target object is identified, performing three-dimensional reconstruction on a scene where the front and top images are located; step S4: during three-dimensional reconstruction, multi-view matching is carried out on a target object; step S5: and generating a three-dimensional reconstruction of the target object to complete positioning. The invention can accurately identify the position of the steel coil, reduce the laser scanning period and reduce the equipment cost by calibrating the binocular camera system, acquiring and identifying the front and top images of the target object by the camera and three-dimensionally reconstructing the front and top scenes.
Description
Technical Field
The invention relates to the technical field of automatic control, in particular to a system, a method, a device and a medium for recognizing object coordinates based on stereoscopic vision, and particularly relates to a system, a method, a device and a medium for recognizing object coordinates based on two paths of binocular stereoscopic vision.
Background
At present, most of steel enterprises adopt an operation mode formed by manual control in the loading, unloading and carrying processes of steel coils in a warehouse. When a person operates a travelling crane, a travelling crane driver or a ground commander needs to visually judge the position of the steel coil or the position of a saddle of the vehicle so as to facilitate the driver to control the travelling crane to travel to a proper position and operate lifting appliances such as a clamp and the like to lift and take the steel coil. When the driver visually observes the steel coil or the saddle position, the driver can judge the steel coil or the saddle position difficultly due to the fact that the driver is higher in position of the driving cab, the steel coil is damaged by the lifting appliance due to misjudgment of the steel coil position easily, or the steel coil is not accurately judged according to the saddle position, so that the steel coil is not accurately positioned on the saddle and rolls. The operation efficiency is affected by long judgment and alignment time, and the steel coil is damaged by friction or rolling, which results in economic loss. If the steel coil or the saddle position is visually observed by a finger hanging worker on the ground, the finger hanging worker station is close to the steel coil or the saddle position, so that the collision between a crane sling or a lifted steel coil and the finger hanging worker is easily caused due to the operation error or equipment failure of a crane driver, and serious casualties are caused.
With the development of detection technology, control technology and equipment manufacturing level, the application of unmanned full-automatic traveling crane in the steel coil warehouse becomes possible gradually. However, in order to realize automatic driving operation, since no driver or finger hanging worker exists on the ground or in the driving cab, the position of the steel coil or the saddle cannot be visually observed manually, and even if a person is arranged, the visual data cannot be converted into computer data and transmitted to a computer system in real time. The accurate identification and positioning of the steel coil are the premise and the basis for accurately hoisting the steel coil by the travelling crane in the warehousing process. Therefore, accurate identification and accurate positioning of the vehicle-mounted steel coil and the vehicle saddle are the primary conditions for automatic loading and unloading and unmanned operation of the steel coil warehouse.
In the warehousing process, the system needs to realize the following work:
1. identification of steel coils in storage: and scanning and imaging the frame vehicle loaded with the steel coil, and distinguishing and identifying the steel coil characteristic data according to the imaging result and the data characteristic.
2. Calculating the coordinate of the steel coil: and calculating three-dimensional physical coordinates (longitudinal, transverse and height) of the central point of the steel coil according to the identification data of the steel coil so as to send and control the travelling crane to be positioned.
3. Calculating the size of the steel coil: and calculating the diameter and width data of the steel coil according to the identification data of the steel coil so as to control the opening width of the clamp of the lifting appliance and the vertical height of the clamp.
At present, a two-dimensional laser scanner is generally adopted and is provided with a suitable high-precision moving mechanism with a third dimension, a three-dimensional scanning function is realized, and the size of a steel coil or a saddle is identified and positioned by adopting filtering and data clustering analysis.
The method adopting the laser three-dimensional scanning modeling can realize the size measurement of the target, but the scanning time is longer, the equipment cost is higher, and the production efficiency is low. Therefore, a method for identifying the coordinates of the steel coil by accurately identifying the position of the steel coil, reducing the laser scanning period and reducing the equipment cost is needed.
Through retrieval, patent document CN105398958A discloses a steel coil measuring, positioning and hoisting method, a device and a crane using the method, wherein the crane is provided with a distance meter, a sensor, a photographing device and a PLC controller. The method comprises the following steps: in the horizontal plane of the working area of the crane, the direction of the head of the carrier for bearing the steel coil to be hung is taken as the X direction, the direction vertical to the X direction is taken as the Y direction, and the vertical direction is taken as the Z direction; and obtaining three-dimensional coordinates (Xo, Yo, Zo) of the central point O of the axis of the steel coil through measurement, photographing, calculation and conversion, and then controlling a crane to hoist the steel coil. In the prior art, the distance measuring instrument and the crane displacement sensor are used as steel coil coordinate measuring equipment, and the steel coil coordinate is calculated by using a curve fitting method, so that the calculated steel coil coordinate is not accurate, the calculated amount is large, and the steel coil assembling efficiency is influenced.
Patent document CN106327044B discloses an automatic identification and positioning device for vehicle-mounted steel coils and vehicle saddles and a method thereof, wherein a wobbler and a linear scanning laser are installed at the top position of a vehicle parking space; the swinging machine is accessed to a first network switch on site through a network cable for communication, and the linear scanning laser is accessed to the first network switch through the network cable; the first network switch is connected with a second network switch in the control machine room through a network cable, and the control terminal and the server are connected to the second network switch so as to realize operation control, data acquisition and three-dimensional positioning calculation of the swinging machine and the linear scanning laser. The server is communicated with the Internet through a third network switch and used for transferring vehicle logistics information and remotely monitoring the operation condition of the server. In the prior art, the steel coil and saddle coordinates are identified by adopting laser scanning and three-dimensional point cloud modeling, the obtained coordinate data are separated and extracted mainly according to the elevation value from the elevation model data, and the steel coil and saddle data are extracted by controlling the upper and lower limit threshold ranges of the elevation value, so that the defects that the upper and lower limit threshold ranges have loading errors and the measurement cannot be accurately carried out are overcome.
Therefore, it is highly desirable to develop a method for accurately recognizing the coordinates of an object in different planes and multiple scenes, and matching the acquired image with the three-dimensional scene information.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide an object coordinate identification system, method, device and medium based on two paths of binocular stereoscopic vision, which can accurately identify the position of a steel coil, reduce the laser scanning period and reduce the equipment cost.
The invention provides a stereoscopic vision-based object coordinate identification method, which comprises the following steps:
step S1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system;
step S2: respectively acquiring front and top images by using a plurality of cameras and identifying a target object based on a neural network;
step S3: after the target object is identified, three-dimensional reconstruction is carried out on the scene where the front and top images are located;
step S4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
step S5: and generating three-dimensional reconstruction of the target object to finish positioning.
Preferably, in step S1, a binocular camera system is used for calibration, that is, an internal parameter matrix and an external parameter matrix are obtained, where the binocular camera system includes an a-path binocular camera and a B-path binocular camera.
Preferably, the three-dimensional reconstruction in step S3 is performed by calibrating and correcting a binocular camera in a binocular camera system and then performing stereo matching, thereby obtaining a disparity map, calculating depth information, and obtaining three-dimensional coordinates of the target object.
Preferably, the a-way binocular camera is used for acquiring a front image of the target object and reconstructing X, Y scenes in the three-dimensional scene, and the B-way binocular camera is used for acquiring a top image of the target object and reconstructing Z scenes in the three-dimensional scene.
Preferably, in step S4, the acquired image and the three-dimensional scene information are acquired from the path a binocular camera and the path B binocular camera, respectively, and the image information is matched to generate three-dimensional coordinate data of the steel coil.
According to the present invention, there is provided a stereoscopic vision-based object coordinate recognition system, comprising:
module M1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system;
module M2: respectively acquiring front and top images by using a plurality of cameras and identifying a target object based on a neural network;
module M3: after the target object is identified, performing three-dimensional reconstruction on a scene where the front and top images are located;
module M4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
module M5: and generating a three-dimensional reconstruction of the target object to complete positioning.
According to the present invention, there is provided a computer readable storage medium having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of the method of any of the above.
According to the present invention, there is provided a stereoscopic vision-based object coordinate recognition apparatus including the stereoscopic vision-based object coordinate recognition system or the computer-readable storage medium storing the computer program.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention can accurately identify the position of the steel coil, reduce the laser scanning period and reduce the equipment cost by calibrating the binocular camera system, acquiring and identifying the front and top images of the target object by the camera and three-dimensionally reconstructing the front and top scenes.
2. According to the invention, the 2-path binocular camera is adopted to respectively identify different planes and X, Y, Z scene coordinates of the steel coil, so that the problem of inaccurate depth scene in the single-path binocular camera three-dimensional scene reconstruction process is solved, and the identification precision in the Z direction in the binocular stereoscopic steel coil coordinate identification system is improved.
3. The object coordinate identification method based on stereoscopic vision not only can be used for identifying the coordinates of the steel coil, but also can be used for identifying the coordinates of objects with regular shapes.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic flow chart of a stereoscopic vision-based object coordinate recognition method according to the present invention;
FIG. 2 is a schematic diagram illustrating the specific steps of the stereoscopic vision-based object coordinate recognition method according to the present invention;
FIG. 3 is a schematic diagram of the position of the test apparatus during calibration according to the present invention;
FIG. 4 is a schematic view of the structure of Mask R-CNN in the present invention;
fig. 5 is a plan layout view of the object coordinate recognition system based on stereoscopic vision in the present invention.
In the figure:
a path A of binocular cameras 1; a crown block 2; a B-path binocular camera 3; a target object 4; a saddle 5.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the invention.
As shown in fig. 1 to 5, the present invention provides a stereoscopic vision-based object coordinate recognition method, including the following steps:
step S1: and calibrating the camera system, and establishing the relation between a multi-path camera pixel coordinate system and a world coordinate system.
Step S1.1: setting the height, distance and calibration range of the camera;
step S1.2: shooting a group of calibration rod images vertical to the ground at certain intervals in the transverse direction and the longitudinal direction;
step S1.3: reading two-dimensional pixel coordinates (left and right cameras) of the picture by manually extracting a calibration point, and generating a data file by the extracted coordinates;
step S1.4: importing a database file into an independently compiled program;
step S1.5: and finally, reconstructing corresponding three-dimensional coordinates of all the calibration points, and completing calibration of internal and external parameters of the camera.
Step S2: and respectively acquiring front and top images by using a plurality of cameras and identifying the target object based on the neural network.
Step S2.1: establishing a material and dump truck effective area morphological feature picture material library;
step S2.2: inputting the whole picture in the material library into CNN to obtain a feature map;
step S2.3: inputting the convolution characteristics into the RPN to obtain the characteristic information of the candidate frame;
step S2.4: judging whether the features (RoI Align) extracted from the candidate frame belong to a specific class by using a classifier;
step S2.5: for a candidate frame belonging to a certain characteristic, further adjusting the position of the candidate frame by using a regressor;
step S2.6: and finally, determining the most suitable target identification frame, and displaying the material contour in the picture in a segmentation mode.
Step S3: after the target object is identified, three-dimensional reconstruction is performed on the scene where the front and top images are located. The three-dimensional reconstruction is to perform stereo matching after calibration and correction of a binocular camera in a binocular camera system, thereby obtaining a disparity map, calculating depth information, and obtaining three-dimensional coordinates of a target object. The A-path binocular camera is used for acquiring a front image of a target object and reconstructing X, Y scenes in a three-dimensional scene, and the B-path binocular camera is used for acquiring a top image of the target object and reconstructing a Z scene in the three-dimensional scene. And the coordinate identification host of the target object acquires the acquired image and the three-dimensional scene information from the A-path binocular camera and the B-path binocular camera respectively, matches the image information and generates the three-dimensional coordinate data of the steel coil. The specific steps taking the identification of the steel coil as an example are as follows:
step S3.1: when the steel coil conveyed by the vehicle reaches the parking space, the travelling crane automatically runs above the parking space, and the target object is located at the designated position in the visual field range of the camera;
step S3.2: two paths of cameras shoot pictures of the object;
step S3.3: extracting screen coordinates of a specific point in the object;
step S3.4: extracted coordinate generation data (visual coordinates);
step S3.5: and converting the visual coordinate into a corresponding three-dimensional coordinate by taking the position data of the travelling crane as a coordinate system to obtain the position coordinate data of the steel coil under the travelling crane coordinate system.
Step S4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
step S5: and generating three-dimensional reconstruction of the target object to finish positioning.
In the present inventionPreferred embodiment(s) of the inventionFor further explanation.
Based on the above embodiment, in step S1 of the present invention, to establish the transformation relationship between the camera coordinate system and the world coordinate system, the calibration of the internal and external parameters of the binocular camera is first performed; the traditional binocular vision adopts a black and white chess grid calibration method to calibrate the internal and external parameters of the camera, and the method is simple to operate and high in precision. However, the distribution range of the angular points in the black and white chess grids is too small compared with the large scene of the unmanned warehouse environment, and the calibration requirement on the camera is difficult to meet, so that the calibration rod is applied to replace the traditional black and white grid calibration plate method. And calibrating by adopting a binocular camera system to obtain the internal and external parameter matrixes, wherein the binocular camera system comprises an A path of binocular cameras and a B path of binocular cameras.
Based on the above embodiment, in step S2, the front and top images are acquired by using the a-way binocular camera and the B-way binocular camera, and the target object is identified by using the Mask R-CNN feature extraction algorithm based on the deep learning frame PyTorch.
Based on the above embodiment, the target object can be identified based on the convolutional neural network in step S2 in the present invention.
The invention provides an object coordinate recognition system based on stereoscopic vision, which comprises:
module M1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system;
module M2: respectively acquiring front and top images by using a plurality of cameras and identifying a target object based on a neural network;
module M3: after the target object is identified, performing three-dimensional reconstruction on a scene where the front and top images are located;
module M4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
module M5: and generating three-dimensional reconstruction of the target object to finish positioning.
The invention also provides a computer-readable storage medium having a computer program stored thereon, which, when being executed by a processor, carries out the steps of the method of any of the above.
The present invention also provides a stereoscopic-vision-based object coordinate recognition apparatus including the stereoscopic-vision-based object coordinate recognition system described above or the computer-readable storage medium described above in which the computer program is stored.
Further, the object coordinate recognition device based on the stereoscopic vision comprises an A path of binocular camera 1, an overhead traveling crane 2, a B path of binocular camera 3, a target object 4 and a saddle 5, wherein the A path of binocular camera 1 is used for acquiring a front image of the target object and reconstructing X, Y scenes in a three-dimensional scene, and the B path of binocular camera 3 is used for acquiring a top image of the target object and reconstructing a Z scene in the three-dimensional scene; overhead traveling crane 2 sets up in B way binocular camera 3 top, and target object 4 places in saddle 5.
Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, units provided by the present invention as pure computer readable program code, the system and its various devices, modules, units provided by the present invention can be fully implemented by logically programming method steps in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units included in the system for realizing various functions can also be regarded as structures in the hardware component; means, modules, units for realizing various functions can also be regarded as structures in both software modules and hardware components for realizing the methods.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (8)
1. An object coordinate identification method based on stereoscopic vision is characterized by comprising the following steps:
step S1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system;
step S2: respectively acquiring front and top images by using a plurality of cameras and identifying a target object based on a neural network;
step S3: after the target object is identified, three-dimensional reconstruction is carried out on the scene where the front and top images are located;
step S4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
step S5: and generating three-dimensional reconstruction of the target object to finish positioning.
2. The object coordinate recognition method based on stereoscopic vision according to claim 1, wherein in step S1, an inside and outside parameter matrix is obtained by calibrating with a binocular camera system, wherein the binocular camera system comprises an a-way binocular camera and a B-way binocular camera.
3. The stereoscopic vision-based object coordinate recognition method of claim 2, wherein the three-dimensional reconstruction in step S3 is performed by calibrating and correcting a binocular camera in a binocular camera system and then performing stereo matching, thereby obtaining a disparity map, calculating depth information, and obtaining three-dimensional coordinates of the target object.
4. The stereoscopic vision-based object coordinate recognition method of claim 2, wherein the a-way binocular camera is used for acquiring a front image of the target object and reconstructing X, Y scenes in the three-dimensional scene, and the B-way binocular camera is used for acquiring a top image of the target object and reconstructing Z scenes in the three-dimensional scene.
5. The stereoscopic vision-based object coordinate recognition method as claimed in claim 2, wherein the step S4 acquires the acquired images and three-dimensional scene information from the a-way binocular camera and the B-way binocular camera, respectively, matches the image information, and generates three-dimensional coordinate data of the steel coil.
6. An object coordinate recognition system based on stereoscopic vision, comprising:
module M1: calibrating a camera system, and establishing a relation between a multi-path camera pixel coordinate system and a world coordinate system;
module M2: respectively acquiring front and top images by using a plurality of cameras and identifying a target object based on a neural network;
module M3: after the target object is identified, three-dimensional reconstruction is carried out on the scene where the front and top images are located;
module M4: during three-dimensional reconstruction, multi-view matching is carried out on a target object;
module M5: and generating three-dimensional reconstruction of the target object to finish positioning.
7. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 5.
8. An object coordinate recognition apparatus based on stereoscopic vision, characterized by comprising the object coordinate recognition system based on stereoscopic vision according to claim 6 or a computer-readable storage medium storing a computer program according to claim 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110291250.5A CN115115768A (en) | 2021-03-18 | 2021-03-18 | Object coordinate recognition system, method, device and medium based on stereoscopic vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110291250.5A CN115115768A (en) | 2021-03-18 | 2021-03-18 | Object coordinate recognition system, method, device and medium based on stereoscopic vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115115768A true CN115115768A (en) | 2022-09-27 |
Family
ID=83324459
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110291250.5A Pending CN115115768A (en) | 2021-03-18 | 2021-03-18 | Object coordinate recognition system, method, device and medium based on stereoscopic vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115115768A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115565134A (en) * | 2022-10-13 | 2023-01-03 | 广州国交润万交通信息有限公司 | Ball machine monitoring blind area diagnosis method, system, equipment and storage medium |
| CN117495698A (en) * | 2024-01-02 | 2024-02-02 | 福建卓航特种设备有限公司 | Flying object identification method, system, intelligent terminal and computer readable storage medium |
| CN118015599A (en) * | 2024-04-09 | 2024-05-10 | 深圳市前海铼停科技有限公司 | Binocular parking identification method, system and processor based on image stitching |
-
2021
- 2021-03-18 CN CN202110291250.5A patent/CN115115768A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115565134A (en) * | 2022-10-13 | 2023-01-03 | 广州国交润万交通信息有限公司 | Ball machine monitoring blind area diagnosis method, system, equipment and storage medium |
| CN115565134B (en) * | 2022-10-13 | 2024-03-15 | 广州国交润万交通信息有限公司 | Diagnostic method, system, equipment and storage medium for monitoring blind area of ball machine |
| CN117495698A (en) * | 2024-01-02 | 2024-02-02 | 福建卓航特种设备有限公司 | Flying object identification method, system, intelligent terminal and computer readable storage medium |
| CN118015599A (en) * | 2024-04-09 | 2024-05-10 | 深圳市前海铼停科技有限公司 | Binocular parking identification method, system and processor based on image stitching |
| CN118015599B (en) * | 2024-04-09 | 2024-06-25 | 深圳市前海铼停科技有限公司 | Binocular parking identification method, system and processor based on image stitching |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112476434B (en) | Visual 3D pick-and-place method and system based on cooperative robot | |
| CN115115768A (en) | Object coordinate recognition system, method, device and medium based on stereoscopic vision | |
| CN109685066B (en) | Mine target detection and identification method based on deep convolutional neural network | |
| CN110497187B (en) | Sun flower pattern assembly system based on visual guidance | |
| CN106662451A (en) | Lost vehicle recovery utilizing associated feature pairs | |
| DE102020118635A1 (en) | 3D data generation device and robot control system | |
| CN113173502B (en) | Anticollision method and system based on laser vision fusion and deep learning | |
| CN112950532B (en) | Train pantograph state detection method | |
| CN115131268A (en) | Automatic welding system based on image feature extraction and three-dimensional model matching | |
| CN112528979A (en) | Transformer substation inspection robot obstacle distinguishing method and system | |
| CN115482195A (en) | Train part deformation detection method based on three-dimensional point cloud | |
| CN115880372A (en) | Unified calibration method and system for external hub positioning camera of automatic crane | |
| CN106643661A (en) | Rail-mounted crane suspender pose detection system based on machine vision and method thereof | |
| JPWO2020090897A1 (en) | Position detection device, position detection system, remote control device, remote control system, position detection method, and program | |
| CN111476062A (en) | Lane line detection method and device, electronic equipment and driving system | |
| CN116276938B (en) | Mechanical arm positioning error compensation method and device based on multi-zero visual guidance | |
| CN114202548A (en) | Forklift pallet positioning method and device, storage medium and electronic equipment | |
| CN116309882A (en) | Tray detection and positioning method and system for unmanned forklift application | |
| CN112614191B (en) | Loading and unloading position detection method, device and system based on binocular depth camera | |
| CN115330751A (en) | Bolt detection and positioning method based on YOLOv5 and Realsense | |
| CN103383574B (en) | A kind of high-precision motion compensating alignment system and high-precision motion compensating localization method | |
| CN107316302A (en) | Driving automatic positioning equipment and its corresponding localization method based on binocular vision | |
| CN207115499U (en) | Driving automatic positioning equipment based on binocular vision | |
| CN112466112A (en) | Apparatus and method for generating vehicle data and system | |
| CN113643374A (en) | Multi-view camera calibration method, device, equipment and medium based on road characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |