CN114494644A - Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system - Google Patents
Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system Download PDFInfo
- Publication number
- CN114494644A CN114494644A CN202210066808.4A CN202210066808A CN114494644A CN 114494644 A CN114494644 A CN 114494644A CN 202210066808 A CN202210066808 A CN 202210066808A CN 114494644 A CN114494644 A CN 114494644A
- Authority
- CN
- China
- Prior art keywords
- cooperative target
- binocular
- point cloud
- stereo matching
- spatial non
- 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
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- 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 Vision & Pattern Recognition (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method and a system thereof, wherein the method comprises the following steps: s1, acquiring a binocular image of the space non-cooperative target and correcting the binocular image; s2, carrying out unsupervised depth learning based on the binocular stereo matching depth neural network model, and calculating by using a binocular image of a space non-cooperative target to obtain a disparity map of the space non-cooperative target; s3, calculating spatial non-cooperative target point cloud data; s4, calculating to obtain a spatial non-cooperative target point cloud to be registered, and initializing a spatial non-cooperative target reference point cloud; s5, carrying out registration by utilizing an ICP (inductively coupled plasma) algorithm, and solving a rotation angle and a translation amount; s6, adding new point cloud data in the current binocular image into the reference point cloud to serve as a new reference point cloud in subsequent registration; and S7, adding new point cloud data while calculating the pose, and finishing three-dimensional reconstruction. The method has the beneficial effects that the pose estimation and the three-dimensional reconstruction of the spatial non-cooperative target are accurately carried out.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the technical field of spatial image processing, in particular to a method and a system for estimating and reconstructing a spatial non-cooperative target pose based on binocular stereo matching.
[ background of the invention ]
An iterative near point method ICP (iterative close point) is based on a Point Set To Point Set (PSTPS) registration method, namely a point set to point set registration method based on quaternion, and after a corresponding near point set is determined from a measurement point set, a new near point set is calculated by using a method proposed by Faugera and Hebert; and (5) carrying out iterative calculation until the objective function value formed by the residual sum of squares is unchanged, and ending the iterative process. The ICP algorithm is mainly used for solving the registration problem based on the free form surface. Binocular stereo matching is a binocular vision research hotspot, a binocular camera shoots left and right viewpoint images of the same scene, a disparity map is obtained by using a stereo matching algorithm, and a depth map is further obtained. Binocular stereo matching can be generally divided into four steps: matching cost calculation, cost aggregation, parallax calculation and parallax optimization.
Deep Learning (DL) is a research direction in the field of Machine Learning (ML), which is introduced to make it closer to the original goal-Artificial Intelligence (AI). Deep Neural Networks (DNNs) are the basis of Deep learning, and DNNs are also sometimes called Multi-Layer perceptrons (MLPs), and are divided from DNNs according to the positions of different layers, and the Neural network layers inside DNNs can be divided into three types, an input Layer, a hidden Layer and an output Layer. Supervised learning is a machine learning task that infers functions from a labeled training dataset. Unsupervised learning is the solution of various problems in pattern recognition from training samples whose classes are unknown (not labeled).
In recent years, autonomous on-orbit service of satellites has received increasing attention from various countries, and has become a focus of research. At present, the in-orbit service technology for cooperative spacecraft is relatively mature, but a considerable part of in-orbit service objects belong to non-cooperative targets, such as space debris, fault satellites, rocket upper-level and non-cooperative spacecraft and the like. Limited by the non-cooperative target, complexity and the prior art, no precedent for completely and autonomously realizing the on-orbit service of the non-cooperative spacecraft exists at present, and the autonomous on-orbit service technology for researching the space non-cooperative target has very important significance and application value.
Through retrieval and analysis, the related patents and the defects thereof in the prior art are as follows: the invention discloses a document CN105976353A, the name of which is a space non-cooperative target pose estimation method based on model and point cloud global matching, wherein the pose estimation is carried out by using an ICP (inductively coupled plasma) algorithm, but the defects are that a depth camera is required to be used, a known target model point cloud is required, and the method is not possible under the condition of a non-cooperative target and does not comprise the function of carrying out three-dimensional reconstruction on the non-cooperative target; the invention discloses a document CN106679634A, the name of which is 'a space non-cooperative target pose measurement method based on stereoscopic vision', uses stereoscopic vision to estimate the pose, but has the defects that the edge information needs to be extracted, the process is complicated, the available points of three-dimensional reconstruction are few, and the reconstruction model is not accurate; the invention publication CN113313821A entitled "a fast three-dimensional reconstruction system and method" performs three-dimensional reconstruction on a target, but has the disadvantages that a depth camera is required and a plurality of cameras are required.
Aiming at the defects of the prior art, the invention technically improves the pose estimation and the three-dimensional reconstruction of the spatial non-cooperative target.
[ summary of the invention ]
The invention aims to provide a method for accurately estimating the pose of a spatial non-cooperative target and reconstructing the pose in three dimensions.
In order to achieve the purpose, the technical scheme adopted by the invention is a method for estimating and reconstructing the pose of a spatial non-cooperative target based on binocular stereo matching, which comprises the following steps:
s1, acquiring binocular images of the space non-cooperative target, namely left images and right images of the space non-cooperative target, by using binocular cameras, namely a left camera and a right camera, in the process of flying around the non-cooperative target by using the binocular cameras, and correcting the binocular images of the space non-cooperative target by using calibration data of the binocular cameras;
s2, calculating to obtain a spatial non-cooperative target disparity map by using the spatial non-cooperative target binocular image obtained in the step S1 based on the unsupervised deep learning of the binocular stereo matching depth neural network model;
s3, calculating point cloud data of the spatial non-cooperative target by using the calibration parameters of the binocular camera and the spatial non-cooperative target disparity map obtained in the step S2;
s4, calculating to obtain a spatial non-cooperative target point cloud to be registered based on the spatial non-cooperative target point cloud data in the step S3, and initializing a spatial non-cooperative target reference point cloud;
s5, registering the point cloud to be registered of the space non-cooperative target and the reference point cloud of the space non-cooperative target obtained in the step S4 by utilizing an ICP (inductively coupled plasma) algorithm, acquiring the binocular image pose of the current space non-cooperative target, and calculating the rotation angle and the translation amount of the space non-cooperative target from the pose transformation matrix of the binocular image of the current space non-cooperative target;
s6, adding new point cloud data in the current space non-cooperative target binocular image in the step S4 to the space non-cooperative target reference point cloud according to the rotation angle and the translation amount of the space non-cooperative target calculated in the step S5, and using the new point cloud data as a new space non-cooperative target reference point cloud in the subsequent step S5;
and S7, repeating the steps S4 to S6, adding new point cloud data into the spatial non-cooperative target reference point cloud while calculating the spatial non-cooperative target pose, and finishing the spatial non-cooperative target three-dimensional reconstruction.
Preferably, the binocular stereo matching-based method for estimating and reconstructing the pose of the spatial non-cooperative target further comprises a binocular stereo matching deep neural network model pre-training step:
s01, creating a binocular stereo matching depth neural network model with binocular image input and disparity map output;
and S02, pre-training the binocular stereo matching depth neural network model by using the spatial non-cooperative target public data set to perform supervised deep learning.
Preferably, the binocular stereo matching-based method for estimating and reconstructing the pose of the spatial non-cooperative target further comprises a binocular stereo matching deep neural network model pre-training step:
s03, training the binocular stereo matching depth neural network model in the using process, so that the binocular stereo matching depth neural network model adapts to the space environment, and the output disparity map is more accurate.
Preferably, in the above method for estimating pose of spatial non-cooperative target and reconstructing three-dimensional based on binocular stereo matching, if there is no point cloud of spatial non-cooperative target: step S4 initializes the point cloud to be registered of the spatial non-cooperative target obtained by calculation as a spatial non-cooperative target reference point cloud.
Preferably, in the above method for estimating a pose of a spatial non-cooperative target and reconstructing a spatial non-cooperative target based on binocular stereo matching, the specific method for calculating point cloud data of the spatial non-cooperative target in step S3 is as follows: setting a binocular camera calibration parameter as a pixel focal length (simple description meaning) fx、fyAnd offset (simple descriptive meaning) cx、cyThe base line (in simple interpretation) of the binocular camera is b, and the parallax (in simple interpretation) corresponding to the point (u, v) in the left image and/or the right image in the binocular image is disu，vThen, the calculation formula of the coordinates (x, y, z) of the point (u, v) in space is:
still another object of the present invention is to provide a system for accurately estimating and reconstructing the pose of a spatial non-cooperative target.
In order to achieve the above further object, the present invention adopts a technical scheme that a binocular stereo matching based spatial non-cooperative target pose estimation and three-dimensional reconstruction system comprises a binocular camera, a computing resource and a spatial non-cooperative target, wherein the binocular camera comprises a left camera and a right camera; the binocular camera and the computing resources are carried by the spacecraft and operate in a space orbit; the method is used for executing the binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method.
Preferably, the parameters of the left camera and the right camera are identical, and the parameters of the left camera and the right camera are connected with the computing resource and positioned on the left side and the right side of the computing resource.
Preferably, Super 16mm lenses are adopted by the left camera and the right camera, the focal length is 50mm, and the resolution of the collected images is set to 1366 × 768; the distance between the left camera and the right camera is 0.5m, the orientations of the left camera and the right camera are completely consistent, and the left camera and the right camera are perpendicular to a connecting line.
Preferably, the distance between the binocular camera and the space non-cooperative target is 10-100 m.
The method and the system for estimating and reconstructing the pose of the spatial non-cooperative target based on binocular stereo matching have the following beneficial effects: the method has the advantages that 1, dense depth information of a non-cooperative target can be obtained only by using a pair of binocular cameras and without using a depth camera; 2. the system work flow is simple and clear; 3. the method has the advantages that the non-cooperative target model point cloud is not required to be known, and the non-cooperative target is subjected to three-dimensional reconstruction in a mode of gradually fusing the reference point cloud; the method has the following innovation points that 1, the disparity map is obtained through unsupervised deep learning in the space, so that the neural network adapts to the space environment, and the accurate disparity map is obtained; 2. no known non-cooperative target model point clouds are required; 3. and carrying out three-dimensional reconstruction on the non-cooperative target in a mode of gradually fusing the reference point cloud while carrying out pose estimation on the non-cooperative target.
[ description of the drawings ]
Fig. 1 is a schematic structural diagram of a binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system.
FIG. 2 is a flow chart of a binocular stereo matching-based method for estimating and reconstructing a pose of a spatial non-cooperative target.
[ detailed description ] embodiments
The invention is further described with reference to the following examples and with reference to the accompanying drawings.
Example 1
The embodiment realizes a binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system.
Fig. 1 is a schematic structural diagram of a binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system. As shown in fig. 1, the binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system of the present embodiment is composed of binocular cameras (a left camera and a right camera) and computing resources.
The parameters of the left camera and the right camera are completely the same, and the left camera and the right camera are connected with the computing resource and are respectively positioned at the left side and the right side. In a simulation environment, Super 16mm lenses are adopted by the binocular cameras, the focal length is 50mm, the image resolution is set to 1366 × 768, the distance between the left camera and the right camera is 0.5m, the directions of the left camera and the right camera are completely consistent, and the left camera and the right camera are perpendicular to the connecting line of the left camera and the right camera.
In the using process, the distance between the binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction system and the non-cooperative target is 10-100 m; and keeping the camera shooting the non-cooperative target, performing fly-around on the non-cooperative target, and completing pose resolving and three-dimensional reconstruction in the fly-around of the non-cooperative target.
The system for estimating the pose of the spatial non-cooperative target and reconstructing the pose of the spatial non-cooperative target based on binocular stereo matching is usually carried by a spacecraft and operates in an orbit, a binocular camera respectively shoots images of the non-cooperative target such as a fault satellite in a camera view field to obtain a left image and a right image, binocular stereo matching is carried out by utilizing computing resources, and then pose calculation and three-dimensional reconstruction of the non-cooperative target are completed.
Example 2
The embodiment realizes a binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method, and is based on the embodiment 1 of a binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction system.
FIG. 2 is a flow chart of a method for estimating and reconstructing a pose of a spatial non-cooperative target based on binocular stereo matching. As shown in fig. 2, the method for estimating and reconstructing the pose of the spatial non-cooperative target based on binocular stereo matching in the present embodiment includes the following steps:
step 1, acquiring binocular images of a space non-cooperative target by using a binocular camera, and correcting the binocular images by using calibration data of the binocular camera;
step 2, calculating by using the binocular image obtained in the step 1 to obtain a disparity map by using an unsupervised binocular stereo matching algorithm based on deep learning; the specific process is as follows:
step 2-1, creating a depth neural network model, inputting a binocular image, and outputting a disparity map;
step 2-2, designing a supervision loss function, and pre-training the deep neural network established in the step 2-1 by using an open data set;
2-3, designing an unsupervised loss function, using the deep neural network pre-trained in the step 2-2 in a space environment, and enabling the deep neural network to adapt to the space environment in the using process so as to enable an output disparity map to be more accurate;
step 3, calculating to obtain point cloud data of a space non-cooperative target by using the calibration parameters of the binocular camera and the disparity map obtained in the step 2, and using the point cloud data as reference point cloud in the subsequent step; the method for calculating the point cloud data of the space non-cooperative target comprises the following steps:
assuming the camera calibration parameter as the pixel focal length fx、fyAnd offset cx、cyThe base line of the binocular camera is b, and the parallax corresponding to the point (u, v) in the left image (or the right image) in the binocular image is disu，vThen, the calculation formula of the coordinates (x, y, z) of the point (u, v) in space is:
step 4, using the method from the step 1 to the step 3, continuously acquiring binocular images and calculating point clouds to be registered while the binocular camera performs flying around on the non-cooperative target along with the satellite;
step 5, registering the point cloud to be registered in the step 4 and the reference point cloud in the step 3 by utilizing an ICP (inductively coupled plasma) algorithm, acquiring the pose of the current frame, and calculating a rotation angle and a translation amount from a pose transformation matrix;
step 6, adding the point cloud of the current frame in the step 4 into the reference point cloud according to the rotation angle and the translation amount calculated in the step 5, and using the point cloud as the reference point cloud matched with the subsequent point cloud;
and 7, repeating the steps 4 to 6, adding new points into the reference point cloud while calculating the pose of the spatial non-cooperative target, and finishing three-dimensional reconstruction.
The binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method has the advantages that: dense depth information of a non-cooperative target can be obtained only by using a pair of binocular cameras without using a depth camera; the system work flow is simple and clear; the non-cooperative target model point cloud is not required to be known, and the non-cooperative target is subjected to three-dimensional reconstruction in a mode of gradually fusing the reference point cloud.
The embodiment of the invention discloses an innovative point of a space non-cooperative target pose estimation and three-dimensional reconstruction method based on binocular stereo matching, which comprises the following steps: acquiring a disparity map in an unsupervised deep learning space, so that a neural network adapts to a space environment to obtain an accurate disparity map; no known non-cooperative target model point clouds are required; and carrying out three-dimensional reconstruction on the non-cooperative target in a mode of gradually fusing the reference point cloud while carrying out pose estimation on the non-cooperative target.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a computer readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and additions can be made without departing from the principle of the present invention, and these should also be considered as the protection scope of the present invention.
Claims (9)
1. A binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method is characterized by comprising the following steps:
s1, acquiring binocular images of the space non-cooperative target, namely left images and right images of the space non-cooperative target, by using binocular cameras, namely a left camera and a right camera, in the process of flying around the non-cooperative target by using the binocular cameras, and correcting the binocular images of the space non-cooperative target by using calibration data of the binocular cameras;
s2, calculating to obtain a spatial non-cooperative target disparity map by using the spatial non-cooperative target binocular image obtained in the step S1 based on the unsupervised deep learning of the binocular stereo matching depth neural network model;
s3, calculating point cloud data of the spatial non-cooperative target by using the calibration parameters of the binocular camera and the spatial non-cooperative target disparity map obtained in the step S2;
s4, calculating to obtain a spatial non-cooperative target point cloud to be registered based on the spatial non-cooperative target point cloud data in the step S3, and initializing a spatial non-cooperative target reference point cloud;
s5, registering the point cloud to be registered of the space non-cooperative target and the reference point cloud of the space non-cooperative target obtained in the step S4 by utilizing an ICP (inductively coupled plasma) algorithm, acquiring the binocular image pose of the current space non-cooperative target, and calculating the rotation angle and the translation amount of the space non-cooperative target from the pose transformation matrix of the binocular image of the current space non-cooperative target;
s6, adding new point cloud data in the current space non-cooperative target binocular image in the step S4 to the space non-cooperative target reference point cloud according to the rotation angle and the translation amount of the space non-cooperative target calculated in the step S5, and using the new point cloud data as a new space non-cooperative target reference point cloud in the subsequent step S5;
and S7, repeating the steps S4 to S6, adding new point cloud data into the spatial non-cooperative target reference point cloud while calculating the spatial non-cooperative target pose, and finishing the spatial non-cooperative target three-dimensional reconstruction.
2. The binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method according to claim 1, characterized by further comprising a binocular stereo matching deep neural network model pre-training step:
s01, creating a binocular stereo matching depth neural network model with binocular image input and disparity map output;
and S02, pre-training the binocular stereo matching depth neural network model by using the spatial non-cooperative target public data set to perform supervised deep learning.
3. The binocular stereo matching-based space non-cooperative target pose estimation and three-dimensional reconstruction method according to claim 2, characterized by further comprising a binocular stereo matching deep neural network model pre-training step:
s03, training the binocular stereo matching depth neural network model in the using process, so that the binocular stereo matching depth neural network model adapts to the space environment, and the output disparity map is more accurate.
4. The binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method according to claim 1, wherein if no spatial non-cooperative target reference point cloud exists: step S4 initializes the point cloud to be registered of the spatial non-cooperative target obtained by calculation as a spatial non-cooperative target reference point cloud.
5. The method of claim 1, wherein the method comprises a binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction methodThe method is characterized in that the specific method for calculating the point cloud data of the spatial non-cooperative target in the step S3 is as follows: setting a calibration parameter of a binocular camera as a pixel focal length fx、fyAnd offset cx、cyThe base line of the binocular camera is b, and the parallax corresponding to the point (u, v) in the left image and/or the right image in the binocular image is disu，vThen, the calculation formula of the coordinates (x, y, z) of the point (u, v) in space is:
6. a binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system is characterized in that: the system comprises binocular cameras, computing resources and a spatial non-cooperative target, wherein the binocular cameras comprise a left camera and a right camera; the binocular camera and the computing resources are carried by the spacecraft and operate in a space orbit; the method is used for executing the binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method as claimed in any one of claims 1 to 5.
7. The binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system according to claim 6, wherein: the parameters of the left camera and the right camera are completely the same, are connected with the computing resource and are positioned on the left side and the right side of the computing resource.
8. The binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system according to claim 7, wherein: the left camera and the right camera adopt Super 16mm lenses, the focal length is 50mm, and the resolution of the collected images is set to 1366 x 768; the distance between the left camera and the right camera is 0.5m, the orientations of the left camera and the right camera are completely consistent, and the left camera and the right camera are perpendicular to a connecting line.
9. The binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction system according to claim 8, wherein: the distance between the binocular camera and the space non-cooperative target is 10-100 m.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210066808.4A CN114494644A (en) | 2022-01-20 | 2022-01-20 | Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210066808.4A CN114494644A (en) | 2022-01-20 | 2022-01-20 | Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114494644A true CN114494644A (en) | 2022-05-13 |
Family
ID=81472197
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210066808.4A Pending CN114494644A (en) | 2022-01-20 | 2022-01-20 | Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114494644A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116363205A (en) * | 2023-03-30 | 2023-06-30 | 中国科学院西安光学精密机械研究所 | Space target pose resolving method based on deep learning and computer program product |
| CN117115228A (en) * | 2023-10-23 | 2023-11-24 | 广东工业大学 | SOP chip pin coplanarity detection method and device |
-
2022
- 2022-01-20 CN CN202210066808.4A patent/CN114494644A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116363205A (en) * | 2023-03-30 | 2023-06-30 | 中国科学院西安光学精密机械研究所 | Space target pose resolving method based on deep learning and computer program product |
| CN117115228A (en) * | 2023-10-23 | 2023-11-24 | 广东工业大学 | SOP chip pin coplanarity detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111862126B (en) | Non-cooperative target relative pose estimation method combining deep learning and geometric algorithm | |
| Wang et al. | 360sd-net: 360 stereo depth estimation with learnable cost volume | |
| CN112258618B (en) | Semantic mapping and positioning method based on fusion of prior laser point cloud and depth map | |
| CN112927357B (en) | 3D object reconstruction method based on dynamic graph network | |
| CN111563415B (en) | Binocular vision-based three-dimensional target detection system and method | |
| CN112102458A (en) | Single-lens three-dimensional image reconstruction method based on laser radar point cloud data assistance | |
| CN110176032B (en) | Three-dimensional reconstruction method and device | |
| CN110223377A (en) | One kind being based on stereo visual system high accuracy three-dimensional method for reconstructing | |
| JP2021518622A (en) | Self-location estimation, mapping, and network training | |
| CN111998862B (en) | BNN-based dense binocular SLAM method | |
| CN114494644A (en) | Binocular stereo matching-based spatial non-cooperative target pose estimation and three-dimensional reconstruction method and system | |
| Zhu et al. | Latitude: Robotic global localization with truncated dynamic low-pass filter in city-scale nerf | |
| CN113838191A (en) | Three-dimensional reconstruction method based on attention mechanism and monocular multi-view | |
| Alcantarilla et al. | Large-scale dense 3D reconstruction from stereo imagery | |
| CN116097307A (en) | Image processing method and related equipment | |
| KR20220014678A (en) | Method and apparatus for estimating depth of images | |
| CN113313740B (en) | Disparity map and surface normal vector joint learning method based on plane continuity | |
| CN114494395A (en) | Depth map generation method, device and equipment based on plane prior and storage medium | |
| Mharolkar et al. | RGBDTCalibNet: End-to-end online extrinsic calibration between a 3D LiDAR, an RGB camera and a thermal camera | |
| CN117523100A (en) | Three-dimensional scene reconstruction method and device based on neural network and multi-view consistency | |
| CN117197333A (en) | Space target reconstruction and pose estimation method and system based on multi-view vision | |
| Li et al. | MODE: Multi-view omnidirectional depth estimation with 360∘ cameras | |
| Jia et al. | DispNet based stereo matching for planetary scene depth estimation using remote sensing images | |
| Tang et al. | Encoder-decoder structure with the feature pyramid for depth estimation from a single image | |
| CN116182894A (en) | Monocular vision odometer method, device and system and storage medium |
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 |