CN114092553A - Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm - Google Patents
Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm Download PDFInfo
- Publication number
- CN114092553A CN114092553A CN202111342724.0A CN202111342724A CN114092553A CN 114092553 A CN114092553 A CN 114092553A CN 202111342724 A CN202111342724 A CN 202111342724A CN 114092553 A CN114092553 A CN 114092553A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- attitude
- transformation matrix
- icp
- fpfh
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
- G06T7/75—Determining position or orientation of objects or cameras using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Abstract
The application relates to a disordered grabbing attitude estimation method based on an FPFH (field programmable gate hydrodynamics) and ICP (inductively coupled plasma) improved algorithm, which can be used for three-dimensional object identification and attitude estimation in an industrial field based on a three-dimensional attitude estimation method of the FPFH and ICP (finite field hydrodynamics) algorithm in disordered grabbing, solves the problems of object attitude estimation precision and instantaneity in the disordered grabbing process, adopts the FPFH algorithm to carry out object identification and three-dimensional attitude estimation, has rotation invariance and robustness on noise, and can well inhibit the influence of industrial field noise. The method adopts an ICP iterative closest point algorithm, aims at the problems that certain noise, shielding and reflection exist in an industrial scene, workpiece point cloud distribution is uneven and holes exist, modifies the target function of ICP, enables registration to be more accurate and accordingly obtains a more accurate attitude transformation matrix.
Description
Technical Field
The disclosure relates to the technical field of three-dimensional attitude estimation, in particular to a disordered grabbing attitude estimation method, a disordered grabbing attitude estimation device, a control system and a readable storage medium based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithms.
Background
At present, there are many research achievements on object recognition and pose estimation based on images, but the imaging process of a two-dimensional image is mapped from a three-dimensional space to a two-dimensional space, and a large amount of information is lost in the process; also, a non-negligible fact is that the best visual system should be oriented to the three-dimensional world. With the improvement of automation degree, the requirements of the fields of robot navigation, industrial part detection and grabbing and the like on a computer vision system are higher and higher, so that the object recognition and posture estimation based on the two-dimensional image cannot meet the requirements of human beings. At present, the three-dimensional point cloud data is acquired very quickly, and meanwhile, the acquisition of the three-dimensional point cloud data is not influenced by illumination, so that the problems of illumination, posture and the like encountered by a two-dimensional image are avoided, and therefore, people pay attention to the three-dimensional object identification and posture estimation based on the point cloud data.
In the process of disordered grabbing, a posture estimation method which is stable and quick and meets certain precision based on an FPFH (fast Fourier transform and inductively coupled plasma) and an ICP (inductively coupled plasma) improved algorithm is required to be found. Conventional three-dimensional pose estimation is generally divided into two stages: a coarse matching stage and a fine registration stage. Traditional object recognition and three-dimensional pose estimation rough matching methods, such as a Point Feature histogram PFH (Radu Bogdan Rusu, Nico Blow, Zoltan Caba Marton, et al. aligning Point Cloud Views using a periodic Feature histogram [ C ]//2008IEEE/RSJ International Conference on Intelligent Robots and Systems, September 22-26, 2008, Acropolis Convention Center, Nice, France. IEEE, 2008.) proposed by Rusu, map to a histogram to obtain statistical information by calculating features such as a vector angle and a distance between any two vertices, and solve a spatial transformation relationship. The algorithm is long in time consumption and cannot meet the real-time requirement. Subsequently, Tombari proposes SHOT (Tombari F, Salti S, Di Stefano L. Unit signatures of historical scores for local surface description [ C ]// European conference on computer vision. Springer, Berlin, Heidelberg, 2010:356-369.), which improves the calculation efficiency compared with PFH, but has low object pose estimation accuracy and real-time performance, and is not suitable for three-dimensional pose estimation of unordered capture in the industry at the present stage.
Disclosure of Invention
In view of the above, the present disclosure provides a method and a system for estimating a posture of unordered grabbing based on an FPFH and an ICP improved algorithm, which solve the problems of accuracy and instantaneity of estimation of an object posture in an unordered grabbing process.
According to an aspect of the disclosure, a disordered grabbing posture estimation method based on an FPFH and ICP improved algorithm is provided, which includes the following steps:
s100, acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data; and the number of the first and second groups,
s200, processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts(ii) a And the number of the first and second groups,
s300, carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti(ii) a And the number of the first and second groups,
s400, transforming a matrix T according to the rough matching posturesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
In a possible implementation manner, optionally, in step S200, the preprocessed point cloud data is processed by using an FPFH object recognition and three-dimensional attitude estimation algorithm to obtain a coarse matching attitude transformation matrix TsThe method comprises the following steps:
s210, performing multi-thread normal estimation on the scene point cloud and the model point cloud respectively, and extracting point cloud key points respectively through down-sampling; and the number of the first and second groups,
s220, respectively constructing local reference systems on the scene point cloud and the model point cloud, calculating topological features of the point cloud key points in feature calculation radiuses under the corresponding local reference systems, and storing the topological features in a histogram; and the number of the first and second groups,
s230, identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between the scene point cloud and the model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud, enabling the model point cloud to approach to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
At one kind canIn an implementation manner of the present invention, optionally, in step S300, the ICP iterative closest point improvement algorithm is adopted to perform accurate pose estimation on the closest point of the model point cloud in the scene point cloud, so as to obtain a pose transformation matrix TiThe method comprises the following steps:
s310, searching the closest point of the model point cloud in the scene point cloud, calculating a corresponding relation, and constructing a covariance matrix H according to the corresponding relation; and the number of the first and second groups,
s320, solving a rotation matrix R and a translational vector T which minimize the objective function according to the constructed covariance matrix H and an introduced ICP improved objective function; and the number of the first and second groups,
s330, applying the rotation matrix R and the translation vector T to the model point cloud, returning to search the closest point of the model point cloud in the scene point cloud, and iterating according to a preset difference threshold and a preset termination condition to obtain a posture transformation matrix Ti。
In a possible implementation manner, optionally, in step S400, the matrix T is transformed according to the rough matching posesAnd the attitude transformation matrix TiCalculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result, wherein the method comprises the following steps:
s410, obtaining the rough matching posture transformation matrix TsAnd the attitude transformation matrix Ti(ii) a And the number of the first and second groups,
s420, transforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix; and the number of the first and second groups,
and S430, converting the obtained final posture transformation matrix into a form of an Euler angle and a translation vector, and displaying a final posture estimation result according to a conversion result.
According to another aspect of the present disclosure, there is provided an apparatus for implementing the method for estimating unordered grasp attitude based on FPFH and ICP improved algorithms, comprising a point cloud data acquisition module, an FPFH algorithm processing module, an ICP algorithm processing module, and an attitude estimation module, wherein,
the point cloud data acquisition module: the system is used for acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data;
the FPFH algorithm processing module: processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts
The ICP algorithm processing module: the method is used for carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti
The attitude estimation module: for transforming a matrix T according to the coarse matching posesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
In a possible implementation manner, optionally, the FPFH algorithm processing module includes:
a point cloud key point extraction module: the system is used for respectively carrying out multithreading normal estimation on the scene point cloud and the model point cloud and respectively extracting point cloud key points through down sampling;
a topological feature calculation module: the device comprises a local reference system, a histogram calculation system, a local reference system, a local image acquisition system and a local image acquisition system, wherein the local reference system is used for respectively constructing a local reference system on a scene point cloud and a model point cloud, calculating the topological features of key points of the point cloud in a feature calculation radius under the corresponding local reference system, and storing the topological features in the histogram;
a pose transformation module: the method is used for identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between scene point cloud and model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud to enable the model point cloud to be close to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
In a possible implementation manner, optionally, the ICP algorithm processing module includes:
a covariance matrix construction module: the covariance matrix H is constructed according to the corresponding relation;
a solution calculation module: the method comprises the steps of improving an objective function according to the constructed covariance matrix H and introduced ICP, and solving a rotation matrix R and a translational vector T which minimize the objective function;
an iteration module: the rotation matrix R and the translation vector T are applied to the model point cloud, the closest point of the model point cloud in the scene point cloud is searched in a return mode, iteration is carried out according to a preset difference threshold value and a preset termination condition, and a posture transformation matrix T is obtainedi。
In one possible implementation, optionally, the pose estimation module includes:
an estimated data acquisition module: for obtaining the coarse matching attitude transformation matrix TsAnd the attitude transformation matrix Ti；
A calculation module: for transforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix;
a visualization module: and the final attitude transformation matrix is used for converting the obtained final attitude transformation matrix into the forms of Euler angles and translation vectors and displaying a final attitude estimation result according to a conversion result.
According to another aspect of the present disclosure, there is also provided a control system including:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to implement the FPFH and ICP improved algorithm-based out-of-order grab pose estimation method when executing the executable instructions.
According to another aspect of the present disclosure, there is also provided a non-transitory computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method for estimating an unordered grasp attitude based on FPFH and ICP improvement algorithms.
The technical effects of this application:
the inventionThe method comprises the steps of obtaining point cloud data of the surface of a workpiece and preprocessing the point cloud data; and processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts(ii) a And carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti(ii) a And transforming a matrix T according to the coarse matching posturesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result. The method can be used for three-dimensional attitude estimation and object pose estimation in the disordered grabbing process based on the FPFH (fast point feature histogram) and ICP (inductively coupled plasma) algorithms, solves the problems of object pose estimation precision and instantaneity in the disordered grabbing process, adopts the FPFH algorithm (fast point feature histogram) object identification and three-dimensional attitude estimation method, has rotation invariance and robustness on noise, and can well inhibit the influence of industrial field noise. In order to further improve the pose estimation precision, a fine registration algorithm needs to be applied after an object identification and three-dimensional pose estimation coarse registration method, and the method adopts (ICP) iterative closest point algorithm, modifies the objective function of ICP, enables registration to be more accurate and obtains a more accurate pose transformation matrix aiming at the problems that certain noise, shielding and reflection exist in an industrial scene and workpiece point cloud distribution is uneven and holes exist easily.
Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.
FIG. 1 is a schematic diagram illustrating an implementation flow of the disordered grab attitude estimation method based on the FPFH and ICP improved algorithms;
FIG. 2 is a schematic view showing the visualization of a recognition result;
FIG. 3 shows a comparison of pose estimation accuracy for the original ICP algorithm and the ICP algorithm incorporating the improved objective function;
fig. 4 shows a comparison diagram of the pose estimation accuracy by the FPFH algorithm and the improved ICP algorithm of the present invention.
Detailed Description
Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. In the drawings, like reference numbers can indicate functionally identical or similar elements. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.
Furthermore, in the following detailed description, numerous specific details are set forth in order to provide a better understanding of the present disclosure. It will be understood by those skilled in the art that the present disclosure may be practiced without some of these specific details. In some instances, methods, means, elements and circuits that are well known to those skilled in the art have not been described in detail so as not to obscure the present disclosure.
According to the method, a good initial position is provided for ICP fine registration by using a FPFH (field programmable gate flash) rough matching method, and a local reference system is constructed by constructing an M matrix and decomposing characteristic values in an FPFH algorithm; calculating topological features under the corresponding reference system, storing the results in a histogram, extracting the features corresponding to the model and the scene point cloud by using a random sampling consistency algorithm, eliminating error point pairs from the features, and resolving to obtain a final rough registration attitude estimation result; searching the closest point of an ICP algorithm, constructing a covariance matrix and solving a rotation matrix and a translation vector which minimize an objective function; according to the problems of uneven distribution and holes of the point cloud of the workpiece caused by noise, shielding and reflection in an industrial scene, the target function is improved, and the problem that the registration precision is reduced when the traditional ICP is applied to the point cloud with the holes collected due to the reflection of the workpiece is solved. The FPFH algorithm has high robustness to noise, rotation invariance and strong identification capability, and can provide a stable and good initial position for the ICP algorithm; the ICP algorithm further improves the pose estimation precision, so that the system can meet the precision requirement in the grabbing task, and the pose estimation precision and the pose estimation real-time performance are improved through interaction of the pose estimation precision and the pose estimation real-time performance.
When the method is used, the calculation radius of a local coordinate system and a feature descriptor is adjusted according to different characteristics of different workpieces in an industrial field, the interpolation interval of a Hough parameter space is adjusted, and the running speed of the method can meet the real-time requirement.
Example 1
As shown in fig. 1, according to an aspect of the present disclosure, there is provided a method for estimating a chaotic grasp attitude based on an FPFH and an ICP improvement algorithm, including the following steps:
s100, acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data;
the manner of acquiring the point cloud data on the surface of the workpiece may be unlimited.
In the embodiment, a photon surface structured light three-dimensional camera is preferentially adopted to acquire point cloud data of the surface of a workpiece (and simultaneously acquire a corresponding two-dimensional image).
After acquisition, because the surface of an industrial field workpiece is smooth, shielding, reflection and the like exist, and a large hole exists in the workpiece point cloud, the acquired point cloud data needs to be preprocessed for point cloud filtering. As shown in the visual schematic diagram of the primary recognition result shown in fig. 2, the object to be detected is a workpiece of an automobile engine, and in a main view window (a large window on the left) in the diagram, the white point cloud is an effect of transforming the model point cloud into a scene by applying a final posture transformation matrix; the black and gray contents are scene point clouds; the initial model point cloud (i.e. the point cloud which is not processed by the attitude transformation matrix) is displayed in the model view window (the small window on the right), the surface of the industrial field workpiece is smooth, shielding, reflection and the like exist, and a large hole exists in the workpiece point cloud. The window also comprises a model loading button, a recognition test button, a disordered grabbing button, a stopping button and an operation button, and the model point cloud of the workpiece to be tested is loaded through the model loading button; the recognition test is used for testing whether the workpiece to be tested can be successfully recognized or not so as to complete the three-dimensional attitude estimation; and (4) disordered grabbing, namely controlling the manipulator to carry out corresponding grabbing actions according to the recognition test result.
The method comprises the following steps of (1) acquiring two-dimensional images of workpieces by using a surface structured light camera (5000 pieces of each type of workpiece are selected in the embodiment) to be assembled and assembled on a production line, filtering most backgrounds, reducing background interference and reducing calculation amount; on the basis of not influencing the point cloud representation capability, point cloud is down-sampled, and the operation amount is further reduced.
S200, processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts；
And (2) processing the preprocessed point cloud data by using an FPFH (field programmable gate hydrographic) object recognition and three-dimensional attitude estimation algorithm, respectively performing multi-thread normal estimation on scene and model point clouds (the model point clouds are generated by software simulation or the scene point clouds completely intercept a workpiece part to form the model point clouds), extracting point cloud key points in down-sampling, and selecting a down-sampling scale to ensure that the running time of the algorithm meets the actual application requirement. After sampling, respectively constructing local reference systems on the scene and the model point cloud, calculating topological features of key points in feature calculation radiuses under the corresponding reference systems, and storing the features in a histogram; finally, obtaining a pose transformation relation between the scene and the model point cloud through Hough transformation, applying the transformation relation to the model point cloud to enable the model point cloud to be close to the scene point cloud, and representing a rough matching posture transformation matrix T by a 4 x 4 matrixs。
FPFH has the rotation invariance and has robustness to the noise, can be fine restrain the influence of industrial field noise.
S300, carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti；
In this embodiment, kdtree is adopted to search the closest point of the model point cloud in the scene point cloud, and the corresponding relationship is foundBuilding a covariance matrix H, solving a rotation matrix R and a translational vector T which minimize the target function according to the built covariance matrix H, returning to the closest point of the search model point cloud in the scene point cloud, and performing the next iteration; calculating the mean square error between corresponding points, and if the mean square error is smaller than a threshold value, stopping iteration; actually, when the program is implemented, the termination condition of iteration is to calculate the difference of the rotation angles between the two iterations, and if the difference is smaller than a difference threshold value, the problem is considered to be converged, and the iteration is terminated; the obtained attitude transformation matrix Ti。
The ICP iterative closest point improvement algorithm can modify the objective function of ICP to enable registration to be more accurate and obtain a more accurate attitude transformation matrix aiming at the problems that certain noise, shielding and reflection exist in an industrial scene and workpiece point cloud distribution is uneven and holes exist easily. As shown in fig. 3, the accuracy of the ICP algorithm is compared to the original ICP algorithm with the improved objective function introduced, with the abscissa being the number of points surrounding the hole and the ordinate being the mean square error. In the experiment in the figure, holes with different sizes are artificially manufactured in the point cloud of the workpiece under the same workpiece and the same scene, and the mean square error (RMSE) of the original ICP algorithm and the improved ICP algorithm when the holes with different sizes exist in the point cloud of the workpiece is compared on the basis, so that the experiment in the figure is realized. Therefore, with the increase of the holes, the improved ICP algorithm is remarkably improved in registration accuracy compared with the original ICP algorithm, is more suitable for industrial sites, and is smaller in mean square error than the original ICP algorithm no matter whether the holes are large or small.
S400, transforming a matrix T according to the rough matching posturesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
Transforming the obtained attitude into matrix TsAnd attitude transformation matrix TiMultiplying to obtain a final attitude transformation matrix; and converting the obtained attitude transformation matrix into the forms of Euler angles and translation vectors, namely the final attitude estimation result.
The FPFH algorithm has high robustness to noise, rotation invariance and strong identification capability; a stable and good initial position can be provided for the ICP algorithm; the ICP algorithm further improves the pose estimation precision, so that the system can meet the precision requirement in the grabbing task.
As shown in fig. 4, in the experiment in the figure, in the industrial field, the pose of the workpiece is randomly adjusted each time, the point cloud of the workpiece is collected, and the point cloud preprocessing is performed, so as to compare the pose estimation accuracy of the FPFH algorithm with the improved ICP algorithm. According to the characteristics of the workpiece (such as the size of a concave or convex part), adjusting the local reference system construction radius, the feature calculation radius and the iteration number of the FPFH feature descriptor; setting the threshold value of the rotation angle difference between two iterations of the ICP algorithm as 1 × e-8The maximum number of iterations is limited to 50. On the basis, the position and attitude of the workpiece are estimated by adopting an FPFH (field programmable gate array) object identification and initial registration algorithm and an ICP (inductively coupled plasma) fine registration algorithm. Carrying out 100 independent experiments, and calculating the average value of the corresponding distances between the model point cloud and the target point cloud (scene point cloud) each time to be used as the standard of measurement errors; the average accuracy of 100 experiments was calculated to observe the overall effect.
As shown in fig. 4, it can be seen that ICP always reduces the error in the case where the FPFH primary registration algorithm provides an initial pose; the abscissa indicates the number of experiments and the ordinate indicates the accuracy in mm. The ICP error reduction effect is particularly obvious under the condition of large initial pose error. The average precision of FPFH initial registration is 2.59543mm, the average precision (mean ICP) of ICP fine registration is 2.26341mm, and due to the fact that shooting positions and angles are different, the model point cloud and the scene point cloud are often greatly different, and actual errors are smaller than those obtained through experiments. In the experiment, the attitude estimation problem can be converged and can reach higher precision in 100 experiments, and the algorithm is proved to have stronger robustness and higher precision and can meet the general unordered grabbing requirement. In this embodiment, a notebook computer of an associated rescuer, i5 for seven generations, 2.5GHz, and the detection time of one recognition result is 1200 ms.
In a possible implementation manner, optionally, in step S200, the preprocessed point cloud data is processed by using an FPFH object recognition and three-dimensional attitude estimation algorithmLine processing to obtain coarse matching posture transformation matrix TsThe method comprises the following steps:
s210, performing multi-thread normal estimation on the scene point cloud and the model point cloud respectively, and extracting point cloud key points respectively through down-sampling;
performing multi-thread normal estimation on a scene and a model point cloud (the model point cloud is generated by software simulation or a workpiece is completely intercepted and selected on the scene point cloud to form the model point cloud), performing down-sampling to extract point cloud key points, and selecting a down-sampling scale to ensure that the running time of the algorithm meets the requirements of practical application; in order to improve the identification success rate, the number of scene point cloud key points is kept to be 3 to 4 times of the number of model point cloud key points.
S220, respectively constructing local reference systems on the scene point cloud and the model point cloud, calculating topological features of the point cloud key points in feature calculation radiuses under the corresponding local reference systems, and storing the topological features in a histogram;
respectively constructing local reference systems on the scene and the model point cloud, calculating topological features of key points in feature calculation radiuses under the corresponding reference systems, and storing the features in a histogram; selecting a local reference system construction radius r, acquiring the size of a part (such as a recess, a bulge and a hole) of the workpiece, which is different from a plane, and keeping the construction radius consistent with the size;
4/3 setting the feature calculation radius as the reference frame construction radius;
for each key point p, and a sphere S with p as the center and radius rpPoint q inside, calculate:
ds＝||p-q||2
M＝0,
M＝M+(r-ds)(p-q)(p-q)T
and (3) carrying out eigenvalue decomposition on the matrix M:
M＝VDV-1,V＝[x+y+z+]
constructing a local reference system from the determined eigenvectors, wherein V is a matrix formed by three eigenvectors of the matrix M, x+，y+，z+Is to take a ball body SpThe three feature vectors with the direction of high point density as the positive direction correspond to the x, y and z axes of the local reference system respectively.
In a local reference system, in a characteristic calculation radius, dividing into 32 regions according to 8 azimuths, 2 pitches and 2 radial directions; and (3) calculating the coordinate of the point q under the local reference system, calculating the included angle theta of a vector from the origin of the coordinate system to the point, simultaneously carrying out four-linear interpolation on the pitch, the radial direction and the azimuth angle of the point at the edge and the Euclidean distance from the origin, and storing the result in a histogram.
S230, identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between the scene point cloud and the model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud, enabling the model point cloud to approach to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
Carrying out object identification by Hough transform, determining a correct feature matching point pair, solving a pose transformation relation between a scene and a model point cloud according to the feature matching point pair, and applying the transformation relation to the model point cloud to enable the model point cloud to be close to the scene point cloud; using 4 x 4 matrices TsAnd (3) representing a coarse matching posture transformation matrix, wherein a 3 multiplied by 3 matrix at the upper sitting corner of the coarse matching posture transformation matrix is a rotation matrix, the first three elements of the last column are translation vectors, the fourth element is 1, and the first three elements of the last row are zero. The interpolation interval of the Hough parameter space is set to be consistent with the model feature point extraction downsampling size, and the Hough voting threshold is set to be 5.0. And adjusting a local coordinate system and a feature calculation radius according to different characteristics of different workpieces in an industrial field, and adjusting region division intervals of Hough parameters.
In a possible implementation manner, optionally, in step S300, the ICP iterative closest point improvement algorithm is adopted to perform accurate pose estimation on the closest point of the model point cloud in the scene point cloud, so as to obtain a pose transformation matrix TiThe method comprises the following steps:
s310, searching the closest point of the model point cloud in the scene point cloud, calculating a corresponding relation, and constructing a covariance matrix H according to the corresponding relation;
searching the closest point of the model point cloud in the scene point cloud by using the kdtree, and constructing a covariance matrix H according to the found corresponding relation;
let p beiRepresenting a point in the model point cloud, qiRepresenting a point in a scene point cloud, qiIs piThen:
wherein the content of the first and second substances,
is the central point of the model point cloud,
the center point of the scene point cloud (possibly a part of points) corresponding to the model is shown, N is the total number of the point pairs, and the matrix H is the constructed covariance matrix.
S320, solving a rotation matrix R and a translational vector T which minimize the objective function according to the constructed covariance matrix H and an introduced ICP improved objective function;
solving a rotation matrix R and a translation which minimize the objective function according to the constructed covariance matrix H
Order to
Then the targetThe minimization problem of the function can be translated into:
(1) finding rotation matrix R-Eq ∑2Minimization;
(2) then the translation vector T is qi-Rpi；
For solving (1), there are available methods of SVD, unit quaternion, dual quaternion, linear least squares, nonlinear least squares, where SVD is taken as an example to solve:
singular value decomposition is performed on the covariance matrix H constructed in step S310:
H＝UΛVT，
wherein U is a left singular vector, V is a right singular vector, the main diagonal of Λ is a singular value, and the other elements are all 0.
Then
R＝VUT，
Apply solution (2) again:
T＝qi-Rpi，
the rotation matrix R and the translation vector T are the result of minimizing the objective function in the current iteration.
Using 4 x 4 matrices TiAnd (3) representing a fine registration posture transformation matrix, wherein a 3 x 3 matrix at the upper sitting corner of the fine registration posture transformation matrix is a rotation matrix R, the first three elements of the last column are translation vectors T, the fourth element is 1, and the first three elements of the last row are zero.
S330, applying the rotation matrix R and the translation vector T to the model point cloud, returning to search the closest point of the model point cloud in the scene point cloud, and iterating according to a preset difference threshold and a preset termination condition to obtain a posture transformation matrix Ti。
Applying the obtained transformation relation (namely the rotation matrix R and the translation matrix T) to the model point cloud, returning to search the closest point of the model point cloud in the scene point cloud, and performing the next iteration; calculating the mean square error between corresponding points, and if the mean square error is smaller than a threshold value, stopping iteration; in factWhen the program is implemented, the termination condition of iteration is to calculate the difference value of the rotation angles between the two iterations, if the difference value is less than the difference threshold value, the problem is considered to be converged, and the iteration is terminated; the obtained attitude transformation matrix Ti(ii) a In an embodiment, the difference threshold value is 1 × e-8. If the difference value is larger than the difference value threshold value, returning to the step 3-1 to continue iteration until a termination condition is met.
The preset difference threshold value and the preset termination condition are selected and set according to actual requirements.
In a possible implementation manner, optionally, in step S400, the matrix T is transformed according to the rough matching posesAnd the attitude transformation matrix TiCalculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result, wherein the method comprises the following steps:
s410, obtaining the rough matching posture transformation matrix TsAnd the attitude transformation matrix Ti(ii) a And the number of the first and second groups,
s420, transforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix; and the number of the first and second groups,
and S430, converting the obtained final posture transformation matrix into a form of an Euler angle and a translation vector, and displaying a final posture estimation result according to a conversion result.
Transforming the attitude obtained in the second step into a matrix TsAnd the attitude transformation matrix T obtained in the third stepiMultiplying to obtain a final attitude transformation matrix; and converting the obtained attitude transformation matrix into the forms of Euler angles and translation vectors, namely the final attitude estimation result.
It should be noted that, although the method of searching the closest point of the model point cloud in the scene point cloud and the hough transform are described above by taking kdtree as an example, those skilled in the art will understand that the present disclosure should not be limited thereto. In fact, the user can flexibly set the method according to personal preference and/or actual application scene, as long as the corresponding result is obtained.
Thus, a good initial position is provided for ICP fine registration by an FPFH rough matching method, and a local reference system is constructed by constructing an M matrix and decomposing characteristic values in an FPFH algorithm; calculating topological features under the corresponding reference system, storing the results in a histogram, extracting the features corresponding to the model and the scene point cloud by using a random sampling consistency algorithm, eliminating error point pairs from the features, and resolving to obtain a final rough registration attitude estimation result; searching the closest point of an ICP algorithm, constructing a covariance matrix and solving a rotation matrix and a translation vector which minimize an objective function; according to the problems of uneven distribution and holes of the point cloud of the workpiece caused by noise, shielding and reflection in an industrial scene, the target function is improved, and the problem that the registration precision is reduced when the traditional ICP is applied to the point cloud with the holes collected due to the reflection of the workpiece is solved. The FPFH algorithm has high robustness to noise, rotation invariance and strong identification capability, and can provide a stable and good initial position for the ICP algorithm; the ICP algorithm further improves the pose estimation precision, so that the system can meet the precision requirement in the grabbing task, and the pose estimation precision and the pose estimation real-time performance are improved through interaction of the pose estimation precision and the pose estimation real-time performance.
Example 2
In the implementation principle of embodiment 1, this embodiment correspondingly provides a device for implementing the method, where the functions and implementation principles of each module of the device are specifically described in embodiment 1, and redundant description is not repeated here.
According to another aspect of the present disclosure, there is provided an apparatus for implementing the method for estimating unordered grasp attitude based on FPFH and ICP improved algorithms, comprising a point cloud data acquisition module, an FPFH algorithm processing module, an ICP algorithm processing module, and an attitude estimation module, wherein,
the point cloud data acquisition module: the system is used for acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data;
the FPFH algorithm processing module: processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts
The ICP algorithmA processing module: the method is used for carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti
The attitude estimation module: for transforming a matrix T according to the coarse matching posesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
In this embodiment, the point cloud data acquisition module preferably selects a photon surface structured light three-dimensional camera, and acquires the point cloud data of the workpiece surface by using the photon surface structured light three-dimensional camera. The method comprises the following steps of (1) acquiring two-dimensional images of workpieces by using a surface structured light camera (5000 pieces of each type of workpiece are selected in the embodiment) to be assembled and assembled on a production line, filtering most backgrounds, reducing background interference and reducing calculation amount; on the basis of not influencing the point cloud representation capability, point cloud is down-sampled, and the operation amount is further reduced.
In a possible implementation manner, optionally, the FPFH algorithm processing module includes:
a point cloud key point extraction module: the system is used for respectively carrying out multithreading normal estimation on the scene point cloud and the model point cloud and respectively extracting point cloud key points through down sampling;
a topological feature calculation module: the device comprises a local reference system, a histogram calculation system, a local reference system, a local image acquisition system and a local image acquisition system, wherein the local reference system is used for respectively constructing a local reference system on a scene point cloud and a model point cloud, calculating the topological features of key points of the point cloud in a feature calculation radius under the corresponding local reference system, and storing the topological features in the histogram;
a pose transformation module: the method is used for identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between scene point cloud and model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud to enable the model point cloud to be close to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
In a possible implementation manner, optionally, the ICP algorithm processing module includes:
a covariance matrix construction module: the covariance matrix H is constructed according to the corresponding relation;
a solution calculation module: the method comprises the steps of improving an objective function according to the constructed covariance matrix H and introduced ICP, and solving a rotation matrix R and a translational vector T which minimize the objective function;
an iteration module: the rotation matrix R and the translation vector T are applied to the model point cloud, the closest point of the model point cloud in the scene point cloud is searched in a return mode, iteration is carried out according to a preset difference threshold value and a preset termination condition, and a posture transformation matrix T is obtainedi。
In one possible implementation, optionally, the pose estimation module includes:
an estimated data acquisition module: for obtaining the coarse matching attitude transformation matrix TsAnd the attitude transformation matrix Ti；
A calculation module: for transforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix;
a visualization module: and the final attitude transformation matrix is used for converting the obtained final attitude transformation matrix into the forms of Euler angles and translation vectors and displaying a final attitude estimation result according to a conversion result.
It will be apparent to those skilled in the art that the modules or steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present invention is not limited to any specific combination of hardware and software.
Example 3
Still further, according to another aspect of the present disclosure, there is also provided a control system including:
a processor;
a memory for storing processor-executable instructions; wherein the processor is configured to execute the executable instructions to implement the FPFH and ICP improved algorithm-based unordered grasp pose estimation method of embodiment 1.
The control system of the disclosed embodiments includes a processor and a memory for storing processor-executable instructions. Wherein the processor is configured to execute the executable instructions to implement any one of the above-described methods for chaotic grab pose estimation based on the FPFH and ICP improvement algorithms.
Here, it should be noted that the number of processors may be one or more. Meanwhile, in the control system of the embodiment of the present disclosure, an input device and an output device may be further included. The processor, the memory, the input device, and the output device may be connected by a bus, or may be connected by other means, and are not limited specifically herein.
The memory, which is a computer-readable storage medium, may be used to store software programs, computer-executable programs, and various modules, such as: the program or the module corresponding to the disordered grabbing attitude estimation method based on the FPFH and ICP improved algorithm in the embodiment of the disclosure. The processor executes various functional applications of the control system and data processing by executing software programs or modules stored in the memory.
The input device may be used to receive an input number or signal. Wherein the signal may be a key signal generated in connection with user settings and function control of the device/terminal/server. The output means may comprise a display device such as a display screen.
Example 4
According to another aspect of the present disclosure, there is also provided a non-transitory computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method for estimating an unordered grasp pose based on FPFH and ICP improvement algorithm of embodiment 1.
Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terms used herein were chosen in order to best explain the principles of the embodiments, the practical application, or technical improvements to the techniques in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (10)
1. A disordered grabbing attitude estimation method based on an FPFH (fast Fourier transform) and ICP (inductively coupled plasma) improved algorithm is characterized by comprising the following steps of:
s100, acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data; and the number of the first and second groups,
s200, processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts(ii) a And the number of the first and second groups,
s300, carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti(ii) a And the number of the first and second groups,
s400, transforming a matrix T according to the rough matching posturesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
2. The method for estimating unordered grabbing attitude based on FPFH and ICP improved algorithms as recited in claim 1, wherein in step S200, the preprocessed point cloud data is processed by using FPFH object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix TsThe method comprises the following steps:
s210, performing multi-thread normal estimation on the scene point cloud and the model point cloud respectively, and extracting point cloud key points respectively through down-sampling; and the number of the first and second groups,
s220, respectively constructing local reference systems on the scene point cloud and the model point cloud, calculating topological features of the point cloud key points in feature calculation radiuses under the corresponding local reference systems, and storing the topological features in a histogram; and the number of the first and second groups,
s230, identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between the scene point cloud and the model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud, enabling the model point cloud to approach to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
3. The method for estimating unordered grabbing attitude based on FPFH and ICP improved algorithms as claimed in claim 1 or 2, wherein in step S300, the ICP iterative closest point improved algorithm is adopted to perform accurate attitude estimation on the closest point of the model point cloud in the scene point cloud to obtain an attitude transformation matrix TiThe method comprises the following steps:
s310, searching the closest point of the model point cloud in the scene point cloud, calculating a corresponding relation, and constructing a covariance matrix H according to the corresponding relation; and the number of the first and second groups,
s320, solving a rotation matrix R and a translational vector T which minimize the objective function according to the constructed covariance matrix H and an introduced ICP improved objective function; and the number of the first and second groups,
s330, applying the rotation matrix R and the translation vector T to the model point cloud, returning to search the closest point of the model point cloud in the scene point cloud, and iterating according to a preset difference threshold and a preset termination condition to obtain a posture transformation matrix Ti。
4. The method for chaotic grasp attitude estimation based on FPFH and ICP improved algorithms according to claim 3, wherein in step S400, said transformation matrix T according to said rough matching attitudesAnd the attitude transformation matrix TiComputing and obtainingAnd finally transforming the attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result, wherein the method comprises the following steps:
s410, obtaining the rough matching posture transformation matrix TsAnd the attitude transformation matrix Ti(ii) a And the number of the first and second groups,
s420, transforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix; and the number of the first and second groups,
and S430, converting the obtained final posture transformation matrix into a form of an Euler angle and a translation vector, and displaying a final posture estimation result according to a conversion result.
5. An apparatus for implementing the disordered grabbing attitude estimation method based on the FPFH and ICP improved algorithms in any one of claims 1 to 4, which is characterized by comprising a point cloud data acquisition module, an FPFH algorithm processing module, an ICP algorithm processing module and an attitude estimation module, wherein,
the point cloud data acquisition module: the system is used for acquiring point cloud data of the surface of a workpiece and preprocessing the point cloud data;
the FPFH algorithm processing module: processing the preprocessed point cloud data by adopting an FPFH (field programmable gate flash) object recognition and three-dimensional attitude estimation algorithm to obtain a rough matching attitude transformation matrix Ts
The ICP algorithm processing module: the method is used for carrying out accurate attitude estimation on the closest point of the model point cloud in the scene point cloud by adopting an ICP iterative closest point improvement algorithm to obtain an attitude transformation matrix Ti
The attitude estimation module: for transforming a matrix T according to the coarse matching posesAnd the attitude transformation matrix TiAnd calculating to obtain a final attitude transformation matrix, and performing conversion processing on the final attitude transformation matrix to obtain a final attitude estimation result.
6. The apparatus of claim 5, wherein the FPFH algorithm processing module comprises:
a point cloud key point extraction module: the system is used for respectively carrying out multithreading normal estimation on the scene point cloud and the model point cloud and respectively extracting point cloud key points through down sampling;
a topological feature calculation module: the device comprises a local reference system, a histogram calculation system, a local reference system, a local image acquisition system and a local image acquisition system, wherein the local reference system is used for respectively constructing a local reference system on a scene point cloud and a model point cloud, calculating the topological features of key points of the point cloud in a feature calculation radius under the corresponding local reference system, and storing the topological features in the histogram;
a pose transformation module: the method is used for identifying objects through Hough transform, determining correct feature matching point pairs, solving the pose transformation relation between scene point cloud and model point cloud according to the feature matching point pairs, applying the transformation relation to the model point cloud to enable the model point cloud to be close to the scene point cloud, and constructing and obtaining a rough matching pose transformation matrix Ts。
7. The apparatus according to claim 5 or 6, wherein said ICP algorithm processing module comprises:
a covariance matrix construction module: the covariance matrix H is constructed according to the corresponding relation;
a solution calculation module: the method comprises the steps of improving an objective function according to the constructed covariance matrix H and introduced ICP, and solving a rotation matrix R and a translational vector T which minimize the objective function;
an iteration module: the rotation matrix R and the translation vector T are applied to the model point cloud, the closest point of the model point cloud in the scene point cloud is searched in a return mode, iteration is carried out according to a preset difference threshold value and a preset termination condition, and a posture transformation matrix T is obtainedi。
8. The apparatus of claim 7, wherein the pose estimation module comprises:
an estimated data acquisition module: for obtaining the coarse matching attitude transformation matrix TsAnd the attitude transformation matrix Ti；
A calculation module: forTransforming the obtained attitude transformation matrix TsWith said attitude transformation matrix TiMultiplying and calculating to obtain a final attitude transformation matrix;
a visualization module: and the final attitude transformation matrix is used for converting the obtained final attitude transformation matrix into the forms of Euler angles and translation vectors and displaying a final attitude estimation result according to a conversion result.
9. A control system, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to implement the FPFH and ICP improved algorithm based unordered grasp pose estimation method of any one of claims 1 to 4 when executing the executable instructions.
10. A non-transitory computer readable storage medium having stored thereon computer program instructions, wherein the computer program instructions, when executed by a processor, implement the method of out-of-order grab pose estimation based on FPFH and ICP improvement algorithms of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111342724.0A CN114092553A (en) | 2021-11-12 | 2021-11-12 | Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111342724.0A CN114092553A (en) | 2021-11-12 | 2021-11-12 | Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114092553A true CN114092553A (en) | 2022-02-25 |
Family
ID=80300519
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111342724.0A Pending CN114092553A (en) | 2021-11-12 | 2021-11-12 | Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114092553A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114743259A (en) * | 2022-02-28 | 2022-07-12 | 华中科技大学 | Pose estimation method, pose estimation system, terminal, storage medium and application |
| CN114783068A (en) * | 2022-06-16 | 2022-07-22 | 深圳市信润富联数字科技有限公司 | Gesture recognition method, gesture recognition device, electronic device and storage medium |
-
2021
- 2021-11-12 CN CN202111342724.0A patent/CN114092553A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114743259A (en) * | 2022-02-28 | 2022-07-12 | 华中科技大学 | Pose estimation method, pose estimation system, terminal, storage medium and application |
| CN114783068A (en) * | 2022-06-16 | 2022-07-22 | 深圳市信润富联数字科技有限公司 | Gesture recognition method, gesture recognition device, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111784770B (en) | Three-dimensional attitude estimation method in disordered grabbing based on SHOT and ICP algorithm | |
| CN109685848B (en) | Neural network coordinate transformation method of three-dimensional point cloud and three-dimensional sensor | |
| Sharma | Comparative assessment of techniques for initial pose estimation using monocular vision | |
| US10515259B2 (en) | Method and system for determining 3D object poses and landmark points using surface patches | |
| CN107045631B (en) | Method, device and equipment for detecting human face characteristic points | |
| CN113223091B (en) | Three-dimensional target detection method, three-dimensional target capture device and electronic equipment | |
| CN111738261A (en) | Pose estimation and correction-based disordered target grabbing method for single-image robot | |
| CN109886124B (en) | Non-texture metal part grabbing method based on wire harness description subimage matching | |
| CN114092553A (en) | Disordered grabbing attitude estimation method based on FPFH (fast Fourier transform and inductively coupled plasma) and ICP (inductively coupled plasma) improved algorithm | |
| WO2022017131A1 (en) | Point cloud data processing method and device, and intelligent driving control method and device | |
| JP5063776B2 (en) | Generalized statistical template matching based on geometric transformation | |
| CN108229305B (en) | Method and device for determining outer frame of target object and electronic equipment | |
| CN114556268B (en) | Gesture recognition method and device and storage medium | |
| CN111598946B (en) | Object pose measuring method and device and storage medium | |
| Kanezaki et al. | High-speed 3d object recognition using additive features in a linear subspace | |
| Figueiredo et al. | Robust cylinder detection and pose estimation using 3D point cloud information | |
| Sun et al. | Robotic grasping using semantic segmentation and primitive geometric model based 3D pose estimation | |
| JP4921847B2 (en) | 3D position estimation device for an object | |
| Hashimoto et al. | Current status and future trends on robot vision technology | |
| WO2019015761A1 (en) | Electronic device, system and method for determining the pose of an object | |
| CN110599407B (en) | Human body noise reduction method and system based on multiple TOF cameras in downward inclination angle direction | |
| Grossmann et al. | Fast view-based pose estimation of industrial objects in point clouds using a particle filter with an icp-based motion model | |
| CN111986299A (en) | Point cloud data processing method, device, equipment and storage medium | |
| Barros et al. | Real-time human pose estimation from body-scanned point clouds | |
| Ninomiya et al. | Deep manifold embedding for 3D object pose estimation |
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 |