CN113318913A - Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision - Google Patents

Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision Download PDF

Info

Publication number
CN113318913A
CN113318913A CN202110137897.2A CN202110137897A CN113318913A CN 113318913 A CN113318913 A CN 113318913A CN 202110137897 A CN202110137897 A CN 202110137897A CN 113318913 A CN113318913 A CN 113318913A
Authority
CN
China
Prior art keywords
glue
camera
image
gbr
reflection
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
Application number
CN202110137897.2A
Other languages
Chinese (zh)
Inventor
张之敬
王磊
朱东升
金鑫
李奕豪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN202110137897.2A priority Critical patent/CN113318913A/en
Publication of CN113318913A publication Critical patent/CN113318913A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
    • B05C—APPARATUS FOR APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
    • B05C5/00—Apparatus in which liquid or other fluent material is projected, poured or allowed to flow on to the surface of the work
    • B05C5/02—Apparatus in which liquid or other fluent material is projected, poured or allowed to flow on to the surface of the work the liquid or other fluent material being discharged through an outlet orifice by pressure, e.g. from an outlet device in contact or almost in contact, with the work
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B05—SPRAYING OR ATOMISING IN GENERAL; APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
    • B05C—APPARATUS FOR APPLYING FLUENT MATERIALS TO SURFACES, IN GENERAL
    • B05C11/00—Component parts, details or accessories not specifically provided for in groups B05C1/00 - B05C9/00
    • B05C11/10—Storage, supply or control of liquid or other fluent material; Recovery of excess liquid or other fluent material

Landscapes

  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention discloses a glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision. An uncalibrated photometric stereo vision solution is provided for the three-dimensional reconstruction problem of the glue dots, and a general bas-relief (GBR) parameter is solved by combining a method based on diffuse reflection and a method based on specular reflection. Firstly, photographing pictures of glue points under multiple light sources, then separating diffuse reflection images from mirror reflection images of the glue points, then utilizing a Bidirectional Reflection Distribution Function (BRDF) to consider the semi-vector symmetry in mirror reflection while finding the Lambert reflection maximum value of the diffuse reflection images, combining the diffuse reflection image and the mirror reflection image, finally solving the parameter value of GBR, recovering a surface normal and reconstructing a three-dimensional surface through integration. The invention adopts an integral method to solve the problem of three-dimensional reconstruction of glue points, so that the reconstruction result has higher accuracy. The invention is suitable for a precision assembly system which has high requirements on the gluing process and strict precision requirements on assembly performance.

Description

Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision
Technical Field
The invention belongs to the field of three-dimensional measurement, and particularly relates to a glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision.
Background
Dispensing technology is an important technology in advanced electronics manufacturing industry, and has wide application in the field of micro-assembly technology and integrated circuit equipment. In the dispensing process, the dispensing amount and the three-dimensional shape of the dispensing point have important influence on the assembly performance of the piece to be bonded. Accurate measurement of the glue sites is of great significance to the micro-assembly technique. The existing method for measuring the amount of the glue applied to the point can be divided into a weighing method and a two-dimensional visual on-line measuring method. For the weighing method, if the glue dots are directly weighed and are submerged because the mass of the glue dots is far less than that of the substrate, the glue dots can only be indirectly measured on the microslide actually, the efficiency is low, and the specific shape of the glue dots is not easy to see. The two-dimensional visual measurement method is used for evaluating glue dispensing consistency through online measurement of glue dot areas, and although the efficiency is high, the method is based on the assumption that glue dots are highly consistent or the glue dots are not deformed. In fact, the glue dots are deformed due to the surface condition of the substrate, the vibration of the worktable and the like, so that the two-dimensional visual measurement has larger errors. Therefore, the method which can automatically identify the three-dimensional shape of the glue dot and has high efficiency is found to play an important role in the field of precision assembly.
The classical photometric stereo method uses a camera to shoot more than three images under different lighting conditions, and can calculate the surface normal of the measured object. However, in the practical application process, the light source calibration is more complicated and has certain limitation to the environment, so that the uncalibrated photometric stereo technology appears, that is, the surface of an object is reconstructed under the condition of an unknown light source. There are many existing methods for solving the stereo technology without calibrating the luminosity, but the existing methods assume that the object is lambertian or directly remove the high luminosity of the picture for research. It is not suitable for three-dimensional reconstruction at glue sites. Therefore, solving the uncalibrated photometric stereo problem of the non-Lambert body is particularly important for the three-dimensional reconstruction of the glue dots.
Disclosure of Invention
In order to solve the problems, the invention provides a glue dot three-dimensional reconstruction method based on an uncalibrated luminosity three-dimensional method. The method can carry out three-dimensional reconstruction on the glue points in the glue dispensing process, does not need to calibrate the light source, and prevents the reconstruction result from being influenced after the direction of the light source is changed. And the method has stronger robustness and higher reconstruction precision than other methods.
A glue dot three-dimensional reconstruction method based on an uncalibrated photometric stereo method comprises the following steps:
step one, a camera device is built on a dispensing machine, and at least three LED point light sources are arranged around the camera. And the dispenser, the LED point light source and the camera are in communication connection, and the dispensing process of the dispenser is controlled.
And step two, starting the camera after the glue dispensing is finished, moving the camera to a position where the glue dispensing point can clearly form an image, sequentially lighting the at least three LED point light sources, lighting only one LED at the same time, and starting to collect the image of the glue dispensing point when each LED is lighted.
And step three, after glue dot image sets under different light sources are obtained, separating the glue dot images into a specular reflection image set and a diffuse reflection image set by using a two-color reflection model.
And step four, after obtaining the diffuse reflection image set and the mirror reflection image set of the glue dots, performing GBR transformation on the diffuse reflection image by using a local maximum value method (LDR) based on Lambert reflection to obtain an initial GBR parameter candidate value set P, and obtaining more accurate GBR parameters from the candidate value set P by ensuring the low-rank structure of the mirror BRDF slice so as to ensure the correct recovery of the surface normal. And combining the LDR method based on diffuse reflection and the method based on mirror reflection to solve the GBR ambiguity problem and obtain the GBR transformation matrix. And solving the surface normal vector matrix of the object according to the GBR transformation matrix.
And step five, performing three-dimensional reconstruction on the normal vector matrix by using an integral method to obtain the real height of the surface of the object.
Has the advantages that:
1. the invention provides a glue dot three-dimensional reconstruction method based on an uncalibrated luminosity three-dimensional method, which can accurately reconstruct the three-dimensional form of a glue dot so as to distinguish whether a glue dispenser can meet the gluing requirement of a product and adjust the parameters of the glue dispenser in time.
2. Compared with the prior three-dimensional reconstruction method, the method can be used under the condition of not calibrating the light source, and avoids the complicated link of light source calibration, which is complicated and difficult. And when the position of the glue point is changed, the light source needs to be calibrated again. The method saves the time of light source calibration, and the obtained reconstruction precision is better than that of the calibrated photometric stereo method.
3. According to the invention, a bicolor reflection model is adopted, image modeling is performed as a linear combination of diffuse reflection and specular reflection, and the diffuse reflection performance of glue dots and the specular reflection condition of the glue dots are considered. The method has stronger robustness and can be suitable for three-dimensional reconstruction of a plurality of different glue components. The three-dimensional shape of the colloid which does not meet the Lambert reflection condition can be accurately reconstructed.
Drawings
FIG. 1 is a flow chart of a glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision
FIG. 2 arrangement of a dispensing device
FIG. 3. actual object
FIG. 4. Diffuse reflectance image of an object
FIG. 5 is a specular reflection image of an object
FIG. 6 depth map of an object
FIG. 7 true three-dimensional morphology of an object
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The invention aims to provide a reference basis for the dispensing performance, monitor the three-dimensional shape of a dispensing point in real time and improve the dispensing quality. And performing three-dimensional reconstruction on the glue points by using a camera and a plurality of LED point light sources under the condition of not calibrating the light sources. The three-dimensional shape of the glue point can be accurately identified, so that the glue point effect can be judged. FIG. 1 is a flow chart of a glue dot three-dimensional reconstruction method based on an uncalibrated photometric stereo method, the steps of the reconstruction method are as follows:
step one, setting a device:
a camera is mounted beside the dispenser, and at least three LEDs are arranged around the camera, each LED forming a point light source. And the dispenser, the LED point light source and the camera are in communication connection, so that the dispensing of the dispenser, the on-off of the LED light source and the on-off of the camera are controlled.
Collecting images of glue spots under different light sources:
after the dispensing machine finishes dispensing, the camera is moved to a position where the dispensing point image can be clearly shot. The light sources are in communication with the camera to control the LED point light sources to be sequentially illuminated, and only one light source is illuminated at a time. And after the light source is lightened, the camera takes pictures to sequentially obtain glue dot images under different light sources.
Step three, processing the image set:
the image set is subjected to image preprocessing, and can be processed by a Gaussian filtering method. Then, according to the principle of a two-color reflection model, the image set is separated into a specular reflection image set and a diffuse reflection image set.
The two-color reflection model is a linear combination of diffuse reflection and specular reflection, wherein I is usedd(x) And Iz(x) Representing diffuse and specular reflection. The observed image i (x) is simply expressed as:
I(x)＝Id(x)+Iz(x) (1)
let Λ (x) and Γ (x) represent the chromaticities of the diffuse and specular components, respectively. Equation (15) can be equivalently written as:
I(x)＝mdΛ(x)+mzΓ(x) (2)
wherein m isdAnd diffuse and specular reflectance, respectively, that depend primarily on the imaging geometry. For specular chromaticity, it is generally uniform.
Step four, solving the GBR parameter value of the uncalibrated photometric stereo method:
general bas-relief blur (GBR) problem:
in the classical photometric stereo method, there are the following relations:
I＝NL (3)
suppose that each picture has P pixel points and K pictures with different illumination. Wherein I ∈ RP×KRepresenting the intensity of each pixel of each picture. N is an element of RP×3To representThe dot product of albedo and surface normal vector. L is belonged to R3×KIndicating the light source direction and light source intensity for each picture.
For uncalibrated photometric stereo methods, i.e., where the light source is unknown, the light source information and the surface information can constitute a bilinear constraint problem. N and L are obtained by a GBR transformation matrix G.
I∈NG-1GL (4)
Applying the integrability constraint, ambiguity arises:
Figure BDA0002927465380000051
the method comprises three parameters of mu, nu and lambda, wherein the lambda is generally more than 0, mu and nu belongs to R. Order:
Figure BDA0002927465380000052
wherein
Figure BDA0002927465380000053
And
Figure BDA0002927465380000054
referred to as pseudo-surface matrices, respectively, are pseudo-light source matrices. The uncalibrated photometric stereo problem can be reduced to only 3 parameters, which is the general bas-relief blur. Therefore, to solve the stereo problem of the non-calibrated luminosity, the GBR problem must be solved, that is, three parameters of μ, ν, and λ in the matrix G are solved.
The solution of the GBR problem of the invention is as follows:
the invention adopts a new method to solve the GBR parameter value. Firstly, for a diffuse reflection image set, finding a local gray maximum point by using a local maximum method based on Lambert reflection, solving a candidate set of GBR parameters, and marking the candidate set as P. After the candidate set P is obtained, the mirror reflection image is ensured to recover the low-rank structure of the estimated mirror BRDF slice, and correct values of mu, v and lambda are selected from P, so that the GBR problem is solved.
The specific principle is as follows:
for a diffuse reflectance image, assuming that at any K light source positions, corresponding to p being a local intensity maximum point, it satisfies the reflectance maximum function:
Figure BDA0002927465380000061
this is a form of minimizing the energy function, which can be collated:
Figure BDA0002927465380000062
to solve for the extremum, the above equation is derived by taking λ as a derivative and letting the derivative be 0:
Figure BDA0002927465380000063
the respective derivatives of μ and v are found to be in [ μ [ ]0,ν0]TAnd [ mu ] and1,ν1]Ta line segment in between. Two-dimensional point on this line segment
Figure BDA0002927465380000064
Can be expressed as
Figure BDA0002927465380000065
Figure BDA0002927465380000066
3D point
Figure BDA0002927465380000067
Form a semicircular curve with a radius of
Figure BDA0002927465380000068
The GBR value is theoretically at the intersection of all semi-circular curves. The parameters of mu, v and lambda can be calculated within the error range by using any two seed points. The present invention re-estimates the LDR method in conjunction with the specular reflected image.
For specular reflectance images, analysis was performed using the Bidirectional Reflectance Distribution Function (BRDF) and half-vector symmetry. When there is a sufficient normal to the surface of the object, we can estimate a 2-dimensional slice of the BRDF when the normal n and the ray direction are known. When the surface normal and the direction of light are distorted by the GBR transform, the two-dimensional BRDF slice no longer has a low rank structure.
Two-dimensional slices were represented using BRDF:
f(ωin,ωout,x)＝ρ(x)+fs(ωin,ωout) (12)
where x denotes the surface point, ρ is the scattering albedo, fs(ωin,ωout) Is a specular bi-directional reflection distribution function.
An objective function is defined to determine the extent to which the estimated two-dimensional BRDF slice satisfies a particular low rank constraint.
The objective function is defined as follows:
Figure BDA0002927465380000071
where N is the number of input specular reflection images, a two-dimensional BRDF slice f can be estimated from each imageiInteger Ki,θhIs the ith chip row thetahIs determined.
The value of the objective function formula solved in the candidate set P is the parameter value of μ, ν, λ in the GBR transformation matrix.
After the GBR transformation matrix is obtained, the direction quantity set N of the object can be obtained by using the formula (6).
Step five, obtaining a normal vector
Figure BDA0002927465380000072
And then, performing three-dimensional reconstruction on the normal vector by using an integration method:
because the orientation of the glue dots is known, the ambiguity problem in GBR ambiguity can be resolved. One basis for defining the glue dot is the 0 depth value, integrated in the x and y directions, respectively. Considering a point A (x, y, z (x, y)) of an object in an image, we can define its tangential vector along the x, y coordinate axes as
Figure BDA0002927465380000073
For smooth objects with continuous surfaces, the tangent vector of the point can be approximately represented by the vector from A (x, y) to A (x +1, y)
Figure BDA0002927465380000074
The tangent vectors in both x and y directions can be written approximately as follows:
Figure BDA0002927465380000075
after approximation, the form of the tangent vector of the coordinate representation of any a point with respect to x, y, z can be obtained:
Figure BDA0002927465380000076
according to the fact that the tangent vector is perpendicular to the normal vector, the following results can be obtained:
Figure BDA0002927465380000081
therefore, a depth value of 0 is set for the glue dot picture, and for each pixel of the known normal value in the image, the depth value of the dot can be calculated by using its neighboring pixels. And finally obtaining a three-dimensional depth map of the glue dots.
Examples
The method performs experiments on the real object to achieve the purpose of three-dimensional reconstruction.
Step one, obtaining an object picture under multiple light sources, wherein one object picture is shown in figure three. The picture source is DiLiGenT database.
Step two, preprocessing the picture set and separating the picture set into a mirror reflection picture set I in batches1And diffuse reflection picture set I2. Fig. 4 and 5 are a diffuse reflection image and a specular reflection image, respectively, after separation.
Step three, for I2And carrying out SVD (singular value decomposition) to initialize GBR (global busy range) parameters, searching an LDR (low density resistor) gray maximum value set, calculating a reflection maximum value function and solving a candidate set P of the GBR parameters mu, v and lambda. Then to I1And calculating a low-rank structure of the BRDF slice, and solving the minimum value of the objective function. The solved mu, v and lambda parameter values are the parameter values of the final GBR transformation matrix, and a normal vector matrix N of the object is solved by using a formula. The normal vector result for this object is shown in fig. 6.
And step four, solving the depth value of the object by using an integration method, and representing the height of the object in a gray scale image mode.
The final results are shown in FIG. 7.
Through a plurality of experiments, the average deviation between the reconstruction depth and the true value of the method is 7.39%, which shows that the method has good accuracy.

Claims (6)

1. A glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo is characterized by comprising the following steps:
step one, a glue point identification system is built, a camera is arranged near a glue point machine, and at least 3 LED point light sources are arranged around the camera.
And building a control circuit, wherein the control circuit is in communication connection with the dispenser, the LED point light source and the camera respectively, so that the dispensing process of the dispenser is controlled, and the camera is moved to identify the dispensing points after dispensing is finished.
And step two, after the dispensing machine finishes dispensing, moving the camera to a specified position. The light sources are in communication with the camera to control the LED point light sources to be sequentially illuminated, and only one light source is illuminated at a time. And after the light source is lightened, the camera takes pictures to sequentially obtain glue dot images under different light sources.
And step three, carrying out image preprocessing on the image set, wherein the image preprocessing can be carried out by using a Gaussian filtering method. And then separating the image set into a specular reflection image set and a diffuse reflection image set according to the principle of a two-color reflection model.
And step four, solving the GBR parameter value of the uncalibrated photometric stereo method.
The method specifically comprises the steps of carrying out GBR conversion on a diffuse reflection image based on a Lambert reflection local maximum value method (LDR) to obtain an initial GBR parameter candidate value set P, and finally obtaining more accurate GBR parameters from P by ensuring a low-rank structure of a mirror BRDF slice so as to ensure correct recovery of a surface normal.
And step five, obtaining the height of the glue dots from the surface normal vector by using an integration method.
2. The building glue point identification device according to claim 1, wherein one camera and at least 3 LED point light sources are arranged near a glue point machine and communicate.
3. The building glue dot recognition device according to claim 2, wherein the camera is arranged on a displacement table so that the camera can correctly find the position of the glue dot. The light source is invariant with respect to the camera position.
4. The method of solving for the normal vector of a non-Lambertian object of claim 3, wherein the image is decomposed into a specular reflection image and a diffuse reflection image using a two-color reflection model.
5. The method for solving the uncalibrated photometric stereo according to claim 4 wherein the uncalibrated photometric stereo problem is solved in combination with the specular reflection image and the diffuse reflection image. And for the diffuse reflection image set, finding a local gray maximum value point by using an LDR method, and ensuring the recovery of the low-rank structure of the estimated mirror BRDF slice for the mirror reflection image. The two methods are combined to solve the GBR problem of the uncalibrated photometric stereo method.
6. The method of claim 5, wherein a depth value of 0 is set for the glue dot image, and for each pixel with a known normal value in the image, the depth value of the glue dot can be calculated by using its neighboring pixels, and finally a three-dimensional depth map of the glue dot is obtained.
CN202110137897.2A 2021-02-01 2021-02-01 Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision Pending CN113318913A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110137897.2A CN113318913A (en) 2021-02-01 2021-02-01 Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110137897.2A CN113318913A (en) 2021-02-01 2021-02-01 Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision

Publications (1)

Publication Number Publication Date
CN113318913A true CN113318913A (en) 2021-08-31

Family

ID=77413470

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110137897.2A Pending CN113318913A (en) 2021-02-01 2021-02-01 Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision

Country Status (1)

Country Link
CN (1) CN113318913A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102607455A (en) * 2012-02-27 2012-07-25 华中科技大学 Three-dimensional measurement method for microstructure based on optical microscope and variable illumination
CN103644957A (en) * 2013-12-03 2014-03-19 华中科技大学 Machine vision-based dispensing quality detection method
JP2017121629A (en) * 2017-02-28 2017-07-13 株式会社キーレックス Sealer coating device
CN108982546A (en) * 2018-08-29 2018-12-11 燕山大学 A kind of intelligent robot gluing quality detecting system and method
CN109701812A (en) * 2018-12-27 2019-05-03 北海绩迅电子科技有限公司 A kind of dispenser and the adhesive dispensing method applied to print cartridge
CN112164105A (en) * 2020-08-27 2021-01-01 周波 Method for combining binocular vision with uncalibrated luminosity vision

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102607455A (en) * 2012-02-27 2012-07-25 华中科技大学 Three-dimensional measurement method for microstructure based on optical microscope and variable illumination
CN103644957A (en) * 2013-12-03 2014-03-19 华中科技大学 Machine vision-based dispensing quality detection method
JP2017121629A (en) * 2017-02-28 2017-07-13 株式会社キーレックス Sealer coating device
CN108982546A (en) * 2018-08-29 2018-12-11 燕山大学 A kind of intelligent robot gluing quality detecting system and method
CN109701812A (en) * 2018-12-27 2019-05-03 北海绩迅电子科技有限公司 A kind of dispenser and the adhesive dispensing method applied to print cartridge
CN112164105A (en) * 2020-08-27 2021-01-01 周波 Method for combining binocular vision with uncalibrated luminosity vision

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
赵延松: "未标定光度立体视觉技术研究", 《中国优秀硕士学位论文全文数据库（信息科技辑）》 *

Similar Documents

Publication Publication Date Title
CN109741405B (en) Depth information acquisition system based on dual structured light RGB-D camera
CN101887477B (en) Method for customizing digitalized shoe trees according to a plurality of images of foot shapes
CN103328959A (en) Substrate inspecting apparatus, substrate inspecting system, and method for displaying image for checking substrate inspection results
CN108090896B (en) Wood board flatness detection and machine learning method and device and electronic equipment
CN101907448A (en) Depth measurement method based on binocular three-dimensional vision
CN102243103A (en) Method for quickly measuring colors and three-dimensional profile of object
CN111207670A (en) Line structured light calibration device and method
CN108917640A (en) A kind of laser blind hole depth detection method and its system
Ciortan et al. A practical reflectance transformation imaging pipeline for surface characterization in cultural heritage
CN110487214A (en) A kind of detection system and its detection method of the product qualification rate combined based on photometric stereo and structured light technique
CN113781434A (en) Defect detection method and device, intelligent terminal and computer readable storage medium
CN101719287B (en) Method for rebuilding shape of hemisphere three-dimensional surface with control point information
CN106918300A (en) A kind of large-sized object three-dimensional Measured data connection method based on many three-dimensional tracking devices
CN112097684B (en) Visual measurement system and method based on multi-attitude band structured light
CN116399241B (en) Patch type inductance geometric parameter measurement method and system
CN113318913A (en) Glue dot three-dimensional reconstruction method based on uncalibrated photometric stereo vision
CN110966937B (en) Large member three-dimensional configuration splicing method based on laser vision sensing
CN104655041A (en) Industrial part contour line multi-feature extracting method with additional constraint conditions
CN109540041B (en) Binocular active vision vector orthogonal normalization automobile morphology measurement system and method
DE102007037131A1 (en) Surface i.e. specular reflecting surface, three-dimensional measurement method, involves calculating disparity between synthetic image and real image, which is recorded by one camera, and corrected depth chart
CN112990190A (en) Method and device for identifying reading of pointer instrument
CN108665471A (en) A kind of human body back curve acquisition methods and system based on camera
CN108759715A (en) A kind of method for three-dimensional measurement based on smart mobile phone image
CN115661349B (en) Three-dimensional reconstruction method, system, equipment, medium and product based on sample image
CN110084887B (en) Three-dimensional reconstruction method for space non-cooperative target relative navigation model

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
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210831