CN107564091A - A kind of three-dimensional rebuilding method and device based on quick corresponding point search - Google Patents
A kind of three-dimensional rebuilding method and device based on quick corresponding point search Download PDFInfo
- Publication number
- CN107564091A CN107564091A CN201710617719.3A CN201710617719A CN107564091A CN 107564091 A CN107564091 A CN 107564091A CN 201710617719 A CN201710617719 A CN 201710617719A CN 107564091 A CN107564091 A CN 107564091A
- Authority
- CN
- China
- Prior art keywords
- speckle image
- image
- sub
- speckle
- pixel
- 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
Abstract
The present invention is applied to field of optical measuring technologies, provide a kind of three-dimensional rebuilding method and device that element is searched based on quick corresponding points, speckle image is projected to testee surface, collects the first speckle image and the second speckle image, and it is filtered, goes distortion to handle；To going the first speckle image after distorting and the second speckle image to carry out horizontality correction and synteny correction, whole Pixel-level is carried out to the image after correction using digital picture related algorithm and corresponds to point search, obtains whole Pixel-level corresponding points；And sub-pix optimization is carried out to whole Pixel-level corresponding points using sub-pix corresponding points optimized algorithm, obtain sub-pixel corresponding points；The three-dimensional coordinate of the whole Pixel-level corresponding points is calculated using binocular stereo vision algorithm for reconstructing, so as to obtain the three-dimensional data on testee surface；Method provided by the invention is computationally intensive when solving the problems, such as to search plain corresponding points using traditional digital picture related algorithm, takes, it is achieved thereby that quickly three-dimensional reconstruction target.
Description
Technical Field
The invention belongs to the technical field of optical measurement, and particularly relates to a three-dimensional reconstruction method and device based on fast corresponding point search.
Background
The fields of industrial measurement, cultural relic protection, human body scanning and the like have urgent needs for rapidly acquiring three-dimensional data of the surface of an object. The three-dimensional reconstruction method based on speckle projection is a non-contact rapid optical three-dimensional digital measurement method, only one speckle image needs to be projected, three-dimensional data of the surface of an object can be reconstructed by using a double camera, the environmental interference resistance is strong, the method is very suitable for rapid three-dimensional digital measurement, and particularly has obvious advantages in the aspect of surface three-dimensional digital measurement of moving objects and deformed objects.
The three-dimensional reconstruction method based on speckle projection has the key points that the corresponding points of speckle images collected by a left camera and a right camera are searched and searched by using a digital image correlation algorithm, but the step has large calculation amount, high calculation complexity and time consumption; this also becomes a limiting factor of the speckle projection three-dimensional reconstruction method in real-time fast three-dimensional scanning measurement application. Therefore, the corresponding point searching efficiency is improved, and the method has important significance for improving the time efficiency of three-dimensional reconstruction.
Disclosure of Invention
The invention provides a three-dimensional reconstruction method and a three-dimensional reconstruction device based on rapid corresponding point search, and aims to solve the problems of large calculation amount and time consumption when corresponding points are searched by using a traditional digital image correlation algorithm, so that a rapid three-dimensional reconstruction target is realized.
The invention provides a three-dimensional reconstruction method based on fast corresponding point search, which is applied to a three-dimensional reconstruction system, wherein the three-dimensional reconstruction system comprises: a first imaging device, a projection device, and a second imaging device, the first and second imaging devices being located on either side of the projection device, the method comprising:
projecting a speckle image to the surface of a measured object by using a projection device, and synchronously acquiring the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
respectively carrying out noise point filtering processing on the first speckle image and the second speckle image, and removing imaging distortion of the filtered first speckle image and the filtered second speckle image by using an internal parameter of a pre-calibrated imaging device and a distortion removal formula to obtain a first speckle image and a second speckle image after distortion removal;
performing horizontal correction on the first and second speckle images after being subjected to distortion removal, so that polar lines of the first and second speckle images are adjusted to be parallel to the direction of a u axis of the image to obtain a first and second speckle images after horizontal correction, performing collinear correction on a v coordinate of the other speckle image by taking one of the first and second speckle images after horizontal correction as a reference, so that conjugate polar lines of the first and second speckle images are collinear, so that the polar lines of the first and second speckle images are the same as the intersection point coordinate of the v axis of the image, and obtaining the first and second speckle images after collinear correction;
carrying out integral pixel level corresponding point search on the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain integral pixel level corresponding points;
and performing sub-pixel optimization on the whole-pixel-level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points, and calculating to obtain three-dimensional coordinates of the sub-pixel-level corresponding points by using a binocular stereo vision reconstruction algorithm to obtain three-dimensional data of the surface of the measured object.
Further, the distortion removal formula is as follows:
wherein, the first and the second end of the pipe are connected with each other,
the u 'and the v' are coordinates of the first speckle image and the second speckle image after distortion removal respectively, and the u and the v are coordinates of the first speckle image and the second speckle image which are acquired synchronously by the first imaging device and the second imaging device respectively; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
Further, the formula for performing horizontal correction on the undistorted first speckle image and the second speckle image is as follows:
wherein (u) L ,v L )、(u R ,v R ) Respectively a first speckle image and a second speckle image after distortion removalPoint coordinates of the spot image, (u) L0 ,v L0 )、(u R0 ,v R0 ) Coordinates of poles of image planes of the first and second imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively are image coordinates of corresponding points after the levelness correction;
taking the first speckle image obtained after the horizontal correction as a reference, and carrying out colinearity correction on the v coordinate of the second speckle image according to the following formula:
wherein a, b and c are correction coefficients.
Further, the digital image correlation algorithm is:
wherein, w m To match the half-width size, P, of the correlation search window R (u R ,v R ) Within a sub-window of the second speckle image (u) R ,v R ) The gray-scale value of the point or points,is the mean value of the gray levels of all the points in the second speckle image sub-window, P L (u L ,v L ) Within a sub-window of the first speckle image (u) L vL) the gray value of the point,and omega is a correlation coefficient which is the gray level average value of all points in the first speckle image sub-window.
Further, the performing sub-pixel optimization on the integer-pixel-level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points, and calculating to obtain three-dimensional coordinates of the sub-pixel-level corresponding points by using a binocular stereo vision reconstruction algorithm to obtain three-dimensional data of the surface of the object to be measured, includes:
performing sub-pixel optimization on the whole pixel corresponding point by using a second-order parallax mode optimization algorithm of the first speckle image and the second speckle image after collinearity correction to obtain a sub-pixel level corresponding point coordinate; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein (u) L ,v L )、(u R ,v R ) The coordinates of the corresponding points at the sub-pixel level on the first speckle image and the second speckle image respectively, and the parameter u in the coordinates T Δ v and Δ u are defined as: u. u T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 And assume C L (u L0 ,v L0 ) Is the center point of the sub-image area of the first imaging means, C R (u R0 ,v R0 ) For integer pixel correspondence points, P, corresponding to sub-image areas of the second imaging means L (u L ,v L ) Is C L A point of vicinity, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point;
defining parameters to be optimizedAnd optimizing an objective function:
wherein f is L (u Li ,v Lj )、f R (x Ri ,y Rj ) Is a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second partial derivation;
using the coordinates (u) of the sub-pixel level corresponding points on the obtained first speckle image and the second speckle image L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively obtaining the system parameter matrixes of the first imaging device and the second imaging device which are calibrated in advance.
The invention also provides a three-dimensional reconstruction device based on fast corresponding point search, which is applied to a three-dimensional reconstruction system, and the three-dimensional reconstruction system comprises: a first imaging device, a projection device, and a second imaging device, the first and second imaging devices being located on opposite sides of the projection device, the device comprising:
the acquisition module is used for projecting a speckle image to the surface of a measured object by using the projection device and synchronously acquiring the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
the processing module is used for respectively carrying out noise point filtering processing on the first speckle image and the second speckle image, and removing imaging distortion of the first speckle image and the second speckle image which are subjected to filtering processing by utilizing an internal parameter of a pre-calibrated imaging device and a distortion removal formula to obtain a first speckle image and a second speckle image which are subjected to distortion removal;
the correction module is used for carrying out horizontal correction on the first speckle image and the second speckle image after the distortion removal, enabling polar lines of the first speckle image and the second speckle image to be adjusted to be parallel to the direction of a u axis of the image, obtaining the first speckle image and the second speckle image after the horizontal correction, carrying out co-linear correction on a v coordinate of the other speckle image by taking one of the first speckle image and the second speckle image obtained after the horizontal correction as a reference, enabling conjugate polar lines of the first speckle image and the second speckle image to be collinear, enabling the polar lines of the first speckle image and the second speckle image to be the same as the intersection point coordinate of the v axis of the image, and obtaining the first speckle image and the second speckle image after the co-linear correction;
the searching module is used for searching the corresponding points of the whole pixels of the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain the corresponding points of the whole pixels;
the sub-pixel optimization module is used for performing sub-pixel optimization on the whole-pixel corresponding point by utilizing a sub-pixel corresponding point optimization algorithm to obtain a sub-pixel level corresponding point;
and the reconstruction module is used for calculating to obtain the three-dimensional coordinates of the sub-pixel level corresponding points by utilizing a binocular stereo vision reconstruction algorithm so as to obtain the three-dimensional data of the surface of the measured object.
Further, the distortion removal formula is as follows:
wherein the content of the first and second substances,
the u 'and the v' are coordinates of the first speckle image and the second speckle image after distortion removal respectively, and the u and the v are coordinates of the first speckle image and the second speckle image which are acquired synchronously by the first imaging device and the second imaging device respectively; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
Further, the formula for performing horizontal correction on the undistorted first speckle image and second speckle image is as follows:
wherein (u) L ,v L )、(u R ,v R ) Point coordinates of the first speckle image and the second speckle image after distortion removal, respectively, (u) L0 ,v L0 )、(u R0 ,v R0 ) The coordinates of the poles of the image planes of the first and second imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively are image coordinates of corresponding points after the levelness correction;
taking the first speckle image obtained after the horizontality correction as a reference, and carrying out colinearity correction on the v coordinate of the second speckle image according to the following formula:
wherein a, b and c are correction coefficients.
Further, the digital image correlation algorithm is:
wherein w m To match the half-width size, P, of the correlation search window R (u R ,v R ) Within a sub-window of the second speckle image (u) R ,v R ) The gray-scale value of the point or points,is the mean value of the gray levels of all the points in the second speckle image sub-window, P L (u L ,v L ) Is within a sub-window of the first speckle image (u) L ,v L ) The gray-scale value of the point or points,and omega is a correlation coefficient which is the gray level average value of all points in the first speckle image sub-window.
Further, the sub-pixel optimization module is specifically configured to: performing sub-pixel optimization on the whole pixel corresponding point by using a second-order parallax mode optimization algorithm of the first speckle image and the second speckle image after collinearity correction to obtain a sub-pixel level corresponding point coordinate; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein，(u L ,v L )、(u R ,v R ) The coordinates of the corresponding points at the sub-pixel level on the first speckle image and the second speckle image respectively, and the parameter u in the coordinates T Δ v and Δ u are defined as: u. of T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 And assume C L (u L0 ,v L0 ) Is the center point of the sub-image region of the first imaging device, C R (u R0 ,v R0 ) Is a whole pixel corresponding point, P, corresponding to the sub-image region of the second imaging device L (u L ,v L ) Is C L A point of vicinity, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point;
defining parameters to be optimizedAnd optimizing an objective function:
wherein f is L (u Li ,v Lj )、f R (x Ri ,y Rj ) As a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second order partial derivatives of (d);
the reconstruction module is specifically configured to use the obtained coordinates (u) of the sub-pixel level corresponding points on the first and second speckle images L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively obtaining system parameter matrixes of the first imaging device and the second imaging device through calibration in advance.
Compared with the prior art, the invention has the beneficial effects that: the invention provides a three-dimensional reconstruction method and a device based on fast corresponding point search.A speckle image is projected to the surface of a measured object and collected to obtain a first speckle image and a second speckle image; then, filtering and distortion removing processing are carried out on the obtained first speckle image and the second speckle image; then, performing horizontal correction and collinearity correction on the first speckle image and the second speckle image after distortion removal; searching corresponding points at the whole pixel level on the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain corresponding points at the whole pixel level; performing sub-pixel optimization on the whole pixel level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel level corresponding points; finally, calculating to obtain the three-dimensional coordinates of the integral pixel level corresponding point by using a binocular stereo vision reconstruction algorithm, thereby obtaining the three-dimensional data of the surface of the measured object; compared with the prior art, the method has the advantages that the first speckle image and the second speckle image are subjected to horizontal correction and collinear correction, so that the related search of corresponding points is simplified to the search along the u coordinate axis of the image, the search precision of the corresponding points at the whole pixel level is improved, the search time of the corresponding points at the whole pixel level is reduced, the sub-pixel optimization is carried out on the searched corresponding points at the whole pixel level by utilizing a sub-pixel corresponding point optimization algorithm, the corresponding points at the sub-pixel level are obtained, the number of optimization parameters is greatly reduced, the iterative optimization calculation time is reduced, and the efficiency and the data precision of speckle projection three-dimensional reconstruction are improved.
Drawings
Fig. 1 is a schematic diagram of hardware modules in a three-dimensional reconstruction system according to an embodiment of the present invention;
fig. 2 is a schematic flowchart of a three-dimensional reconstruction method based on fast correspondent point search according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a speckle image projected by a projection apparatus provided by an embodiment of the invention;
FIG. 4-1 is a diagram illustrating a corresponding point search result according to an embodiment of the present invention;
FIG. 4-2 is a schematic diagram of the distribution of the correlation coefficient of the point search along the u-axis according to the embodiment of the present invention;
FIG. 5 is a schematic diagram of a sub-pixel optimized disparity mode according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating the result of scan reconstruction of a human image according to an embodiment of the present invention;
fig. 7 is a block diagram of a three-dimensional reconstruction apparatus based on fast corresponding point search according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In the prior art, the problems of large calculation amount and time consumption when corresponding points are searched by using a traditional digital image correlation algorithm exist, so that the three-dimensional data of the surface of the measured object cannot be quickly acquired.
In order to solve the technical problems, the invention provides a three-dimensional reconstruction method and a device based on fast corresponding point search.A speckle image is projected to the surface of a measured object, a first speckle image and a second speckle image are acquired and are filtered and subjected to distortion removal; then, carrying out horizontal correction and collinearity correction on the first speckle image and the second speckle image after distortion removal; carrying out integral pixel level corresponding point search on the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain integral pixel level corresponding points; then, performing sub-pixel optimization on the whole pixel level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel level corresponding points; and finally, calculating to obtain the three-dimensional coordinates of the integral pixel level corresponding point by using a binocular stereo vision reconstruction algorithm, thereby obtaining the three-dimensional data of the surface of the measured object.
The following specifically describes a three-dimensional reconstruction method based on fast corresponding point search, which is applied to a three-dimensional reconstruction system, as shown in fig. 1, where the three-dimensional reconstruction system includes: a first imaging device 1, a projection device 2 and a second imaging device 3, wherein the first imaging device 1 and the second imaging device 3 are located on two sides of the projection device 2, as shown in fig. 2, the method comprises:
step S101, projecting a speckle image to the surface of a measured object by using a projection device, and synchronously acquiring the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
specifically, before the three-dimensional reconstruction system is used, a first imaging device and a second imaging device of the three-dimensional reconstruction system need to be calibrated, so as to obtain internal parameters and external parameters of the first imaging device and the second imaging device.
Specifically, the first imaging device and the second imaging device described in the embodiment of the present invention are a left camera and a right camera, respectively, and the speckle image shown in fig. 3 is projected onto the surface of the object to be measured by using the projection device, and is synchronously acquired by using the left camera and the right camera, so as to obtain the left speckle image I _ left and the right speckle image I _ right.
Step S102, noise point filtering processing is respectively carried out on the first speckle image and the second speckle image, imaging distortion of the first speckle image and the second speckle image after filtering processing is removed by combining internal parameters of an imaging device calibrated in advance with a distortion removal formula, and the first speckle image and the second speckle image after distortion removal are obtained;
specifically, the distortion removal formula is:
wherein the content of the first and second substances,
u 'and v' are coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) after distortion removal, and u and v are coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) acquired by the first imaging device (left camera) and the second imaging device (right camera) synchronously; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
Step S103, performing horizontal correction on the first and second speckle images after the distortion removal to enable polar lines of the first and second speckle images to be adjusted to be parallel to the direction of a u axis of the image to obtain first and second speckle images after the horizontal correction, performing collinearity correction on a v coordinate of the other speckle image by taking one of the first and second speckle images obtained after the horizontal correction as a reference, and enabling conjugate polar lines of the first and second speckle images to be collinear, so that the polar lines of the first and second speckle images are the same as the intersection coordinate of the v axis of the image, and obtaining the first and second speckle images after the collinearity correction;
specifically, the correction process provided by the embodiment of the present invention is to perform epipolar horizontal correction and collinearity correction on the first speckle image and the second speckle image after distortion removal, respectively.
Wherein, the formula for performing horizontal correction on the first speckle image (left speckle image) and the second speckle image (right speckle image) after the distortion removal is as follows:
wherein (u) L ,v L )、(u R ,v R ) Point coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) after distortion removal, respectively, (u) speckle image L0 ,v L0 )、(u R0 ,v R0 ) Coordinates of the poles of the image planes of the first (left camera) and second (right camera) imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively are image coordinates of corresponding points after the levelness correction;
specifically, the embodiment of the invention takes the first speckle image obtained after the horizontal correction as a reference, and performs the co-linear correction on the v coordinate of the second speckle image; the formula for carrying out the collinearity correction is as follows:
wherein, a, b and c are correction coefficients, namely, the coordinate of the image v of the known corresponding point is substituted into the formula to be obtained by a least square method.
Step S104, searching the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain integral pixel level corresponding points;
specifically, for the first speckle image shown in FIG. 4-1, i.e., the left speckle image, a point P L (u L ,v L ) To (u) L ,v L ) Select (2 w) as center m +1)×(2w m + 1) size image sub-area along v = v on the second, right speckle image L And performing digital correlation search in the direction, and calculating a correlation coefficient between each sub-area of the right speckle image and the sub-area of the left speckle image by using the following formula of a digital image correlation algorithm to obtain the distribution of the correlation coefficients as shown in fig. 4-2.
Specifically, the digital image correlation algorithm is as follows:
wherein, w m To match the half-width size, P, of the correlation search window R (u R ,v R ) In the second, right speckle image sub-window (u) R ,v R ) The gray-scale value of the point or points,is the average of the gray levels of all the points in the sub-window of the right speckle image, P L (u L ,v L ) Is a first speckle image sub-window, namely a left speckle image sub-windowInner (u) L ,v L ) The gray-scale value of the point or points,the average value of the gray levels of all the points in the sub-window of the left speckle image is shown, and omega is a correlation coefficient. If ω is>ω corr (ω corr A grey scale correlation constraint threshold), then the matching pair meets the matching requirement, and then the center point of the search window on the image is the corresponding point.
And S105, performing sub-pixel optimization on the whole-pixel-level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points, and calculating by using a binocular stereo vision reconstruction algorithm to obtain three-dimensional coordinates of the sub-pixel-level corresponding points so as to obtain three-dimensional data of the surface of the measured object.
Specifically, in the embodiment of the present invention, the sub-pixel corresponding point optimization algorithm is an optimization algorithm based on a second-order disparity mode.
Specifically, assume C L (u L0 ,v L0 ) Is the center point, C, of the sub-image area of the first imaging device (left camera) R (u R0 ,v R0 ) To correspond to the integer pixel correspondence point, P, of the sub-image area of the second imaging device (right camera) L (u L ,v L ) Is C L A point nearby, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point; referring to FIG. 5, the parameter u is defined according to the second-order disparity mode T Δ v and Δ u are: u. u T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 。
Specifically, performing sub-pixel optimization on the whole pixel corresponding point by using a second-order parallax mode optimization algorithm of a first speckle image (left speckle image) and a second speckle image (right speckle image) after collinearity correction to obtain sub-pixel level corresponding point coordinates; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein (u) L ,v L )、(u R ,v R ) The coordinates of the sub-pixel level corresponding points on the first speckle image (left speckle image) and the second speckle image (right speckle image), respectively.
Defining parameters to be optimizedAnd optimizing an objective function:
wherein f is L (u Li ,v Lj )、f R (x Ri ,y Rj ) Is a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second order partial derivatives of (d); and circularly iterating according to the formula during solving until the difference value of the results of the previous and subsequent times is smaller than a specified threshold value, and obtaining the optimal solution of the sigma.
Using the coordinates (u) of the sub-pixel level corresponding points on the obtained first speckle image (left speckle image) and second speckle image (right speckle image) L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively calibrating system parameter matrixes of a first imaging device (a left camera) and a second imaging device (a right camera) obtained in advance, wherein the size of the system parameter matrixes is 3 multiplied by 4; fig. 6 shows data obtained by three-dimensional reconstruction of a human image.
The embodiment of the invention provides a three-dimensional reconstruction method based on rapid corresponding point search, which provides an optimized projection correction algorithm and a sub-pixel corresponding point optimization algorithm, greatly improves the efficiency and the precision of corresponding point search time in a speckle three-dimensional reconstruction algorithm, and further improves the efficiency and the data precision of speckle projection three-dimensional reconstruction.
The following describes a three-dimensional reconstruction apparatus based on fast correspondent point search provided by the present invention, where the three-dimensional reconstruction apparatus is applied to a three-dimensional reconstruction system, and the three-dimensional reconstruction system includes: a first imaging device 1, a projection device 2 and a second imaging device 3, wherein the first imaging device 1 and the second imaging device 3 are located at two sides of the projection device 2, as shown in fig. 7, the device comprises:
the collecting module 201 is configured to project a speckle image to a surface of a measured object by using a projecting device, and synchronously collect the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
specifically, before the three-dimensional reconstruction system is used, a first imaging device and a second imaging device of the three-dimensional reconstruction system need to be calibrated, so as to obtain internal parameters and external parameters of the first imaging device and the second imaging device.
Specifically, the first imaging device and the second imaging device described in the embodiment of the present invention are a left camera and a right camera, respectively, the projection device is used to project the speckle image shown in fig. 3 onto the surface of the object to be measured, and the left camera and the right camera are used to perform synchronous acquisition, so as to obtain a left speckle image I _ left and a right speckle image I _ right.
The processing module 202 is configured to perform noise point filtering processing on the first speckle image and the second speckle image, and remove imaging distortion of the filtered first speckle image and the filtered second speckle image by using a pre-calibrated imaging device internal parameter in combination with a distortion removal formula to obtain a first speckle image and a second speckle image after distortion removal;
specifically, the distortion removal formula is:
wherein the content of the first and second substances,
u 'and v' are coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) after distortion removal, and u and v are coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) acquired by the first imaging device (left camera) and the second imaging device (right camera) synchronously; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
The correcting module 203 is used for performing horizontal correction on the first and second speckle images after being subjected to distortion removal, so that polar lines of the first and second speckle images are adjusted to be parallel to the direction of a u axis of the image to obtain the first and second speckle images after being subjected to horizontal correction, performing collinear correction on v coordinates of the other speckle image by taking one of the first and second speckle images after being subjected to horizontal correction as a reference, so that conjugate polar lines of the first and second speckle images are collinear, so that the polar lines of the first and second speckle images are the same as the intersection coordinates of the v axis of the image, and obtaining the first and second speckle images after being subjected to collinear correction;
specifically, the correction process provided by the embodiment of the present invention is to perform epipolar horizontal correction and collinearity correction on the first speckle image and the second speckle image after distortion removal, respectively.
Wherein, the formula for performing horizontal correction on the first speckle image (left speckle image) and the second speckle image (right speckle image) after the distortion removal is as follows:
wherein (u) L ,v L )、(u R ,v R ) (u) point coordinates of the first speckle image (left speckle image) and the second speckle image (right speckle image) after distortion removal, respectively L0 ,v L0 )、(u R0 ,v R0 ) Coordinates of the image plane poles of the first (left camera) and second (right camera) imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively image coordinates of corresponding points after the horizontality correction;
specifically, in the embodiment of the present invention, based on the first speckle image obtained after the horizontal correction, the formula for performing the co-linear correction on the v coordinate of the second speckle image is as follows:
wherein, a, b and c are correction coefficients, namely, the coordinate of the image v of the known corresponding point is substituted into the formula to be obtained by a least square method.
The searching module 204 is configured to perform integer pixel corresponding point searching on the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain an integer pixel corresponding point;
specifically, for the first speckle image, i.e., the left speckle image point P, shown in FIG. 4-1 L (u L ,v L ) In order to (u) L ,v L ) Select (2 w) as center m +1)×(2w m + 1) size image sub-area along v = v on the second, right speckle image L And performing digital correlation search in the direction, and calculating a correlation coefficient between each sub-area of the right speckle image and the sub-area of the left speckle image by using the following formula of a digital image correlation algorithm to obtain the distribution of the correlation coefficients as shown in fig. 4-2.
Specifically, the digital image correlation algorithm is as follows:
wherein, w m To match the half-width size, P, of the correlation search window R (u R ,v R ) In the second speckle image sub-window, i.e. the right speckle image sub-window (u) R ,v R ) The gray-scale value of the point or points,is the average of the gray levels of all the points in the sub-window of the right speckle image, P L (u L ,v L ) Is the first speckle image sub-window, i.e. the left speckle image sub-window (u) L ,v L ) The gray-scale value of the point or points,the average value of the gray levels of all the points in the left speckle image sub-window, and omega is a correlation coefficient. If ω is>ω corr (ω corr A grey scale correlation constraint threshold), then the matching pair meets the matching requirement, and then the center point of the search window on the image is the corresponding point.
A sub-pixel optimization module 205, configured to perform sub-pixel optimization on the integer-pixel corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points;
specifically, in the embodiment of the present invention, the sub-pixel corresponding point optimization algorithm is an optimization algorithm based on a second-order disparity mode.
In particular, assume C L (u L0 ,v L0 ) Is the center point of the sub-image area of the first imaging device (left camera), C R (u R0 ,v R0 ) To the integer pixel correspondence point, P, of the sub-image area of the second imaging device (right camera) L (u L ,v L ) Is C L A point nearby, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point; referring to FIG. 5, the parameter u is defined according to the second-order disparity mode T Δ v and Δ u are: u. of T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 。
Specifically, the sub-pixel optimization module is specifically configured to: performing sub-pixel optimization on the whole pixel corresponding point by using a second-order parallax mode optimization algorithm of the first speckle image (left speckle image) and the second speckle image (right speckle image) after collinearity correction to obtain a sub-pixel level corresponding point coordinate; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein (u) L ,v L )、(u R ,v R ) The coordinates of the sub-pixel level corresponding points on the first speckle image (left speckle image) and the second speckle image (right speckle image), respectively.
Defining parameters to be optimizedAnd optimizing an objective function:
wherein, f L (u Li ,v Lj )、f R (x Ri ,y Rj ) As a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second order partial derivatives of (d); when in solving, the iteration is carried out according to the formula untilAnd if the difference value of the results of the previous and subsequent times is smaller than a specified threshold value, the optimal solution of the sigma can be obtained.
And the reconstruction module 206 is configured to calculate a three-dimensional coordinate of the sub-pixel level corresponding point by using a binocular stereo vision reconstruction algorithm, so as to obtain three-dimensional data of the surface of the object to be measured.
In particular, the reconstruction module is used for utilizing the coordinates (u) of the sub-pixel level corresponding points on the obtained first speckle image (left speckle image) and the second speckle image (right speckle image) L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively calibrating system parameter matrixes of a first imaging device (a left camera) and a second imaging device (a right camera) obtained in advance, wherein the size of the system parameter matrixes is 3 multiplied by 4; fig. 6 shows the data result obtained by three-dimensional reconstruction of a human image.
The three-dimensional reconstruction device based on the rapid corresponding point search provided by the embodiment of the invention provides an optimized projection correction algorithm and a sub-pixel corresponding point optimization algorithm, and greatly improves the efficiency and the precision of corresponding point search time in a speckle three-dimensional reconstruction algorithm, thereby improving the efficiency and the data precision of speckle projection three-dimensional reconstruction.
The above description is intended to be illustrative of the preferred embodiment of the present invention and should not be taken as limiting the invention, but rather, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.
Claims (10)
1. A three-dimensional reconstruction method based on fast correspondent point search is characterized in that the three-dimensional reconstruction method is applied to a three-dimensional reconstruction system, and the three-dimensional reconstruction system comprises: a first imaging device, a projection device, and a second imaging device, the first and second imaging devices being located on either side of the projection device, the method comprising:
projecting a speckle image to the surface of a measured object by using a projection device, and synchronously acquiring the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
respectively carrying out noise point filtering processing on the first speckle image and the second speckle image, and removing imaging distortion of the filtered first speckle image and the filtered second speckle image by using an internal parameter of a pre-calibrated imaging device and a distortion removal formula to obtain a first speckle image and a second speckle image after distortion removal;
performing horizontal correction on the first speckle image and the second speckle image after the distortion removal to enable polar lines of the first speckle image and the second speckle image to be adjusted to be parallel to the direction of a u axis of the image to obtain a first speckle image and a second speckle image after the horizontal correction, performing collinearity correction on a v coordinate of the other speckle image by taking one of the first speckle image and the second speckle image obtained after the horizontal correction as a reference, enabling conjugate polar lines of the first speckle image and the second speckle image to be collinear, enabling the polar lines of the first speckle image and the second speckle image to be the same as the intersection point coordinate of the v axis of the image, and obtaining the first speckle image and the second speckle image after the collinearity correction;
searching corresponding points of the whole pixel level on the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain corresponding points of the whole pixel level;
and performing sub-pixel optimization on the whole-pixel-level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points, and calculating to obtain three-dimensional coordinates of the sub-pixel-level corresponding points by using a binocular stereo vision reconstruction algorithm to obtain three-dimensional data of the surface of the measured object.
2. The three-dimensional reconstruction method of claim 1 wherein the distortion removal formula is:
wherein the content of the first and second substances,
u 'and v' are coordinates of the first speckle image and the second speckle image after distortion removal respectively, and u and v are coordinates of the first speckle image and the second speckle image acquired by the first imaging device and the second imaging device synchronously respectively; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
3. The three-dimensional reconstruction method of claim 1, wherein the formula for performing the horizontal correction on the undistorted first and second speckle images is:
wherein (u) L ,v L )、(u R ,v R ) Point coordinates of the first speckle image and the second speckle image after distortion removal, respectively, (u) L0 ,v L0 )、(u R0 ,v R0 ) The coordinates of the poles of the image planes of the first and second imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively image coordinates of corresponding points after the horizontality correction;
taking the first speckle image obtained after the horizontality correction as a reference, and carrying out colinearity correction on the v coordinate of the second speckle image according to the following formula:
wherein a, b and c are correction coefficients.
4. The three-dimensional reconstruction method of claim 1 wherein said digital image correlation algorithm is:
wherein w m To match the half-width size, P, of the correlation search window R (u R ,v R ) Within a sub-window of the second speckle image (u) R ,v R ) The gray-scale value of the point or points,is the mean value of the gray levels of all the points in the second speckle image sub-window, P L (u L ,v L ) Is within a sub-window of the first speckle image (u) L ,v L ) The gray-scale value of the point or points,and omega is a correlation coefficient which is the gray level average value of all points in the first speckle image sub-window.
5. The three-dimensional reconstruction method according to claim 1, wherein the performing sub-pixel optimization on the whole-pixel-level corresponding points by using a sub-pixel corresponding point optimization algorithm to obtain sub-pixel-level corresponding points, and calculating three-dimensional coordinates of the sub-pixel-level corresponding points by using a binocular stereo vision reconstruction algorithm to obtain three-dimensional data of the surface of the object to be measured comprises:
performing sub-pixel optimization on the whole pixel corresponding point by using a second-order parallax mode optimization algorithm of the first speckle image and the second speckle image after collinearity correction to obtain a sub-pixel level corresponding point coordinate; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein (u) L ,v L )、(u R ,v R ) The coordinates of the corresponding points at the sub-pixel level on the first speckle image and the second speckle image respectively, and the parameter u in the coordinates T Δ v and Δ u are defined as: u. of T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 And assume C L (u L0 ,v L0 ) Is the center point of the sub-image region of the first imaging device, C R (u R0 ,v R0 ) Is a whole pixel corresponding point, P, corresponding to the sub-image region of the second imaging device L (u L ,v L ) Is C L A point nearby, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point;
defining parameters to be optimizedAnd optimizing an objective function:
wherein, f L (u Li ,v Lj )、f R (x Ri ,y Rj ) As a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second partial derivation;
using the coordinates (u) of the sub-pixel level corresponding points on the obtained first speckle image and the second speckle image L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively obtaining the system parameter matrixes of the first imaging device and the second imaging device which are calibrated in advance.
6. A three-dimensional reconstruction device based on fast corresponding point search is characterized in that the three-dimensional reconstruction device is applied to a three-dimensional reconstruction system, and the three-dimensional reconstruction system comprises: a first imaging device, a projection device, and a second imaging device, the first and second imaging devices being located on opposite sides of the projection device, the device comprising:
the acquisition module is used for projecting a speckle image to the surface of a measured object by using the projection device and synchronously acquiring the speckle image of the surface of the measured object by using the first imaging device and the second imaging device to obtain a first speckle image and a second speckle image;
the processing module is used for respectively carrying out noise point filtering processing on the first speckle image and the second speckle image, and removing imaging distortion of the first speckle image and the second speckle image after filtering processing by using pre-calibrated internal parameters of the imaging device and a distortion removal formula to obtain a first speckle image and a second speckle image after distortion removal;
the correction module is used for carrying out horizontal correction on the first speckle image and the second speckle image after the distortion removal, enabling polar lines of the first speckle image and the second speckle image to be adjusted to be parallel to the direction of a u axis of the image, obtaining the first speckle image and the second speckle image after the horizontal correction, carrying out co-linear correction on a v coordinate of the other speckle image by taking one of the first speckle image and the second speckle image obtained after the horizontal correction as a reference, enabling conjugate polar lines of the first speckle image and the second speckle image to be collinear, enabling the polar lines of the first speckle image and the second speckle image to be the same as the intersection point coordinate of the v axis of the image, and obtaining the first speckle image and the second speckle image after the co-linear correction;
the searching module is used for searching the integral pixel corresponding points of the first speckle image and the second speckle image after the collinearity correction by using a digital image correlation algorithm to obtain the integral pixel corresponding points;
the sub-pixel optimization module is used for performing sub-pixel optimization on the whole-pixel corresponding point by utilizing a sub-pixel corresponding point optimization algorithm to obtain a sub-pixel level corresponding point;
and the reconstruction module is used for calculating to obtain the three-dimensional coordinates of the sub-pixel level corresponding points by using a binocular stereo vision reconstruction algorithm so as to obtain the three-dimensional data of the surface of the measured object.
7. The three-dimensional reconstruction apparatus of claim 6 wherein the distortion removal formula is:
wherein the content of the first and second substances,
u 'and v' are coordinates of the first speckle image and the second speckle image after distortion removal respectively, and u and v are coordinates of the first speckle image and the second speckle image acquired by the first imaging device and the second imaging device synchronously respectively; the pre-calibrated internal parameters of the imaging device comprise k 1 ,k 2 ,k 3 ,p 1 ,p 2 ,p 3 ,u 0 ,v 0 Wherein k is 1 ,k 2 ,k 3 Is a pre-calibrated radial distortion coefficient, p 1 ,p 2 ,p 3 Is a pre-calibrated centrifugal distortion coefficient, u 0 、v 0 The image principal point position parameter is calibrated in advance.
8. The three-dimensional reconstruction apparatus of claim 6 wherein the formula for correcting the horizontality of the undistorted first and second speckle images is:
wherein (u) L ,v L )、(u R ,v R ) Point coordinates of the first speckle image and the second speckle image after distortion removal, respectively, (u) L0 ,v L0 )、(u R0 ,v R0 ) Coordinates of poles of image planes of the first and second imaging devices, respectively, (u' L ,v' L )、(u' R ,v' R ) Respectively image coordinates of corresponding points after the horizontality correction;
taking the first speckle image obtained after the horizontal correction as a reference, and carrying out colinearity correction on the v coordinate of the second speckle image according to the following formula:
wherein a, b and c are correction coefficients.
9. The three-dimensional reconstruction apparatus of claim 6 wherein said digital image correlation algorithm is:
wherein, w m To match the half-width size, P, of the correlation search window R (u R ,v R ) Within a sub-window (u) of the second speckle image R ,v R ) The gray-scale value of the point or points,is the mean value of the gray levels of all the points in the second speckle image sub-window, P L (u L ,v L ) Within a sub-window of the first speckle image (u) L ,v L ) The gray-scale value of the point or points,for the first speckle image within the sub-windowThe gray level mean value of all the points, ω, is a correlation coefficient.
10. The three-dimensional reconstruction apparatus of claim 6 wherein said sub-pixel optimization module is specifically configured to: performing sub-pixel optimization on the corresponding points of the whole pixels by using a second-order parallax mode optimization algorithm of the first speckle image and the second speckle image after the collinearity correction to obtain sub-pixel level corresponding point coordinates; wherein, the second-order parallax mode optimization algorithm is as follows:
wherein (u) L ,v L )、(u R ,v R ) The coordinates of the corresponding points at the sub-pixel level on the first speckle image and the second speckle image respectively, and the parameter u in the coordinates T Δ v and Δ u are defined as: u. u T ＝u R0 -u L0 ，Δv＝v L0 -v L ，Δu＝u L -u L0 And assume C L (u L0 ,v L0 ) Is the center point of the sub-image area of the first imaging means, C R (u R0 ,v R0 ) Is a whole pixel corresponding point, P, corresponding to the sub-image region of the second imaging device L (u L ,v L ) Is C L A point of vicinity, P R (u R ,v R ) Is P L (u L ,v L ) A corresponding point;
defining parameters to be optimizedAnd optimizing an objective function:
wherein f is L (u Li ,v Lj )、f R (x Ri ,y Rj ) Is a reference image I L And an image I to be matched R The gray value of each point of the sub-region image,taking the mean value of the gray levels in the region, wherein w is the half width of a window of the sub-region;
each parameter to be optimized is solved by utilizing a Newton-Raphson iterative algorithm, wherein the Newton-Raphson iterative algorithm is as follows:
wherein σ 0 For the parameter value to be optimized corresponding to the coordinate of the corresponding point of the whole pixel,eta (sigma) at sigma 0 The gradient value of the position,Eta (sigma) at sigma 0 Second order partial derivatives of (d);
the reconstruction module is specifically configured to use the obtained coordinates (u) of the sub-pixel level corresponding points on the first and second speckle images L ,v L )、(u R ,v R ) Calculating three-dimensional coordinates (X, Y, Z) of the sub-pixel level corresponding point by combining a binocular stereo vision reconstruction algorithm, thereby obtaining three-dimensional data of the surface of the measured object; the binocular stereo vision reconstruction algorithm comprises the following steps:
wherein, P L 、P R Respectively obtaining system parameter matrixes of the first imaging device and the second imaging device through calibration in advance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710617719.3A CN107564091A (en) | 2017-07-26 | 2017-07-26 | A kind of three-dimensional rebuilding method and device based on quick corresponding point search |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710617719.3A CN107564091A (en) | 2017-07-26 | 2017-07-26 | A kind of three-dimensional rebuilding method and device based on quick corresponding point search |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107564091A true CN107564091A (en) | 2018-01-09 |
Family
ID=60974779
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710617719.3A Pending CN107564091A (en) | 2017-07-26 | 2017-07-26 | A kind of three-dimensional rebuilding method and device based on quick corresponding point search |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107564091A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629841A (en) * | 2018-05-08 | 2018-10-09 | 深圳大学 | One kind being based on laser speckle multiple views three-dimensional vision information method and system |
| CN108734776A (en) * | 2018-05-23 | 2018-11-02 | 四川川大智胜软件股份有限公司 | A kind of three-dimensional facial reconstruction method and equipment based on speckle |
| CN109360246A (en) * | 2018-11-02 | 2019-02-19 | 哈尔滨工业大学 | Stereo vision three-dimensional displacement measurement method based on synchronous sub-district search |
| CN109887022A (en) * | 2019-02-25 | 2019-06-14 | 北京超维度计算科技有限公司 | A kind of characteristic point matching method of binocular depth camera |
| CN110322561A (en) * | 2019-04-30 | 2019-10-11 | 熵智科技(深圳)有限公司 | 3D camera and its measurement method for the unordered sorting of robot |
| CN111243002A (en) * | 2020-01-15 | 2020-06-05 | 中国人民解放军国防科技大学 | Monocular laser speckle projection system calibration and depth estimation method applied to high-precision three-dimensional measurement |
| CN111325683A (en) * | 2020-01-23 | 2020-06-23 | 深圳市易尚展示股份有限公司 | Speckle gray scale correction method and device based on composite coding three-dimensional reconstruction |
| CN113052889A (en) * | 2021-03-24 | 2021-06-29 | 奥比中光科技集团股份有限公司 | Depth calculation method and system |
| CN113256540A (en) * | 2021-07-14 | 2021-08-13 | 智道网联科技(北京)有限公司 | Image distortion removal method and apparatus, electronic device, and computer-readable storage medium |
| FR3107117A1 (en) * | 2020-02-10 | 2021-08-13 | Saint-Gobain Glass France | Method for measuring the geometry of a glazing |
| CN114463251A (en) * | 2021-12-13 | 2022-05-10 | 西安交通大学 | Method and device for measuring deformation of inner surface of intermediate casing of aircraft engine |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875443A (en) * | 2017-01-20 | 2017-06-20 | 深圳大学 | The whole pixel search method and device of the 3-dimensional digital speckle based on grayscale restraint |
-
2017
- 2017-07-26 CN CN201710617719.3A patent/CN107564091A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106875443A (en) * | 2017-01-20 | 2017-06-20 | 深圳大学 | The whole pixel search method and device of the 3-dimensional digital speckle based on grayscale restraint |
Non-Patent Citations (1)
| Title |
|---|
| 石春琴: "随机光照双目立体测量系统中的若干关键问题研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629841A (en) * | 2018-05-08 | 2018-10-09 | 深圳大学 | One kind being based on laser speckle multiple views three-dimensional vision information method and system |
| CN108734776A (en) * | 2018-05-23 | 2018-11-02 | 四川川大智胜软件股份有限公司 | A kind of three-dimensional facial reconstruction method and equipment based on speckle |
| CN108734776B (en) * | 2018-05-23 | 2022-03-25 | 四川川大智胜软件股份有限公司 | Speckle-based three-dimensional face reconstruction method and equipment |
| CN109360246A (en) * | 2018-11-02 | 2019-02-19 | 哈尔滨工业大学 | Stereo vision three-dimensional displacement measurement method based on synchronous sub-district search |
| CN109360246B (en) * | 2018-11-02 | 2021-10-29 | 哈尔滨工业大学 | Stereoscopic vision three-dimensional displacement measurement method based on synchronous subarea search |
| CN109887022A (en) * | 2019-02-25 | 2019-06-14 | 北京超维度计算科技有限公司 | A kind of characteristic point matching method of binocular depth camera |
| CN110322561A (en) * | 2019-04-30 | 2019-10-11 | 熵智科技(深圳)有限公司 | 3D camera and its measurement method for the unordered sorting of robot |
| CN111243002A (en) * | 2020-01-15 | 2020-06-05 | 中国人民解放军国防科技大学 | Monocular laser speckle projection system calibration and depth estimation method applied to high-precision three-dimensional measurement |
| CN111325683A (en) * | 2020-01-23 | 2020-06-23 | 深圳市易尚展示股份有限公司 | Speckle gray scale correction method and device based on composite coding three-dimensional reconstruction |
| CN111325683B (en) * | 2020-01-23 | 2023-06-20 | 深圳市易尚展示股份有限公司 | Speckle gray scale correction method and device based on composite coding three-dimensional reconstruction |
| FR3107117A1 (en) * | 2020-02-10 | 2021-08-13 | Saint-Gobain Glass France | Method for measuring the geometry of a glazing |
| WO2021160618A1 (en) * | 2020-02-10 | 2021-08-19 | Saint-Gobain Glass France | Method for measuring the geometry of a glass panel |
| CN113052889A (en) * | 2021-03-24 | 2021-06-29 | 奥比中光科技集团股份有限公司 | Depth calculation method and system |
| CN113256540A (en) * | 2021-07-14 | 2021-08-13 | 智道网联科技(北京)有限公司 | Image distortion removal method and apparatus, electronic device, and computer-readable storage medium |
| CN113256540B (en) * | 2021-07-14 | 2021-11-19 | 智道网联科技(北京)有限公司 | Image distortion removal method and apparatus, electronic device, and computer-readable storage medium |
| CN114463251A (en) * | 2021-12-13 | 2022-05-10 | 西安交通大学 | Method and device for measuring deformation of inner surface of intermediate casing of aircraft engine |
| CN114463251B (en) * | 2021-12-13 | 2024-03-15 | 西安交通大学 | Method and device for measuring deformation of inner surface of intermediate case of aero-engine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107564091A (en) | A kind of three-dimensional rebuilding method and device based on quick corresponding point search | |
| CN107977997B (en) | Camera self-calibration method combined with laser radar three-dimensional point cloud data | |
| CN106780590B (en) | Method and system for acquiring depth map | |
| CN103003665B (en) | Stereo distance measurement apparatus | |
| US8593524B2 (en) | Calibrating a camera system | |
| US8355564B2 (en) | Corresponding point searching method and three-dimensional position measuring method | |
| CN111080709B (en) | Multispectral stereo camera self-calibration algorithm based on track feature registration | |
| CN110602474B (en) | Method, device and equipment for determining image parallax | |
| CN108648222B (en) | Method and device for improving spatial resolution of structured light depth data | |
| CN105005964A (en) | Video sequence image based method for rapidly generating panorama of geographic scene | |
| WO2018107427A1 (en) | Rapid corresponding point matching method and device for phase-mapping assisted three-dimensional imaging system | |
| CN116433737A (en) | Method and device for registering laser radar point cloud and image and intelligent terminal | |
| CN111462246B (en) | Equipment calibration method of structured light measurement system | |
| CN116957987A (en) | Multi-eye polar line correction method, device, computer equipment and storage medium | |
| CN103914807B (en) | Non-locality image super-resolution method and system for zoom scale compensation | |
| CN108615221B (en) | Light field angle super-resolution method and device based on shearing two-dimensional polar line plan | |
| WO2013134998A1 (en) | 3d reconstruction method and system | |
| WO2018133027A1 (en) | Grayscale constraint-based method and apparatus for integer-pixel search for three-dimensional digital speckle pattern | |
| CN110487254B (en) | Rapid underwater target size measuring method for ROV | |
| CN109990756B (en) | Binocular ranging method and system | |
| CN109373901B (en) | Method for calculating center position of hole on plane | |
| RU2692970C2 (en) | Method of calibration of video sensors of the multispectral system of technical vision | |
| CN110188756B (en) | Product positioning method | |
| CN110631556B (en) | Distance measurement method of heterogeneous stereoscopic vision system | |
| CN112381721A (en) | Human face three-dimensional reconstruction method based on binocular vision |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180109 |