CN114274139A - Automatic spraying method, device, system and storage medium - Google Patents
Automatic spraying method, device, system and storage medium Download PDFInfo
- Publication number
- CN114274139A CN114274139A CN202011037006.8A CN202011037006A CN114274139A CN 114274139 A CN114274139 A CN 114274139A CN 202011037006 A CN202011037006 A CN 202011037006A CN 114274139 A CN114274139 A CN 114274139A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- cloud model
- model
- spraying
- points
- 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.)
- Granted
Links
- 238000005507 spraying Methods 0.000 title claims abstract description 140
- 238000000034 method Methods 0.000 claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims description 41
- 238000013519 translation Methods 0.000 claims description 40
- 230000014616 translation Effects 0.000 claims description 40
- 239000007921 spray Substances 0.000 claims description 28
- 230000008859 change Effects 0.000 claims description 19
- 230000009466 transformation Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 9
- 238000000576 coating method Methods 0.000 claims description 5
- 239000011248 coating agent Substances 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000000844 transformation Methods 0.000 claims description 4
- 239000013598 vector Substances 0.000 description 28
- 230000008569 process Effects 0.000 description 9
- 238000004519 manufacturing process Methods 0.000 description 8
- 238000010422 painting Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000003973 paint Substances 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 230000001131 transforming effect Effects 0.000 description 2
- PUAQLLVFLMYYJJ-UHFFFAOYSA-N 2-aminopropiophenone Chemical compound CC(N)C(=O)C1=CC=CC=C1 PUAQLLVFLMYYJJ-UHFFFAOYSA-N 0.000 description 1
- 101100328518 Caenorhabditis elegans cnt-1 gene Proteins 0.000 description 1
- 208000033748 Device issues Diseases 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 231100000614 poison Toxicity 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000003440 toxic substance Substances 0.000 description 1
Images
Landscapes
- Application Of Or Painting With Fluid Materials (AREA)
- Processing Or Creating Images (AREA)
Abstract
The embodiment of the invention discloses an automatic spraying method, device, system and storage medium. Wherein the method comprises the following steps: generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed; determining a point cloud relationship transformed between the first point cloud model and at least one second point cloud model in a pre-constructed model pool; determining a second point cloud model corresponding to the point cloud relation meeting the preset condition as a point cloud model matched with the first point cloud model; determining a key spraying path point and a spraying track of the first point cloud model based on a pre-configured key spraying path point of a matching point cloud model and the point cloud relation meeting a preset condition; and generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model. According to the technical scheme in the embodiment of the invention, the robot can be automatically sprayed on various products without additional programming.
Description
Technical Field
The invention relates to the technical field of spraying, in particular to an automatic spraying method, device, system and storage medium.
Background
In industrial processes, it is possible to spray paints, coatings, etc. on various products (e.g., furniture, etc.). The spraying mode can be manual spraying by a person or automatic spraying by a robot. Manual spraying can be harmful to the health of workers because workers are exposed to toxic substances and even to explosive substances. The robot automatic spraying can be safer. However, in general, the automatic spraying by robot is to spray a specific product by using a fixed spraying track, and if the product is slightly different or the spraying position or angle is changed, the robot needs to be programmed again, and the automatic spraying mode is not flexible.
Disclosure of Invention
In view of this, an aspect of the embodiments of the present invention provides an automatic spraying method, apparatus, system and storage medium, which can plan a spraying track according to a model in a model pool without performing additional programming on a robot, and perform automatic spraying on a product.
The embodiment of the invention provides an automatic spraying method, which comprises the following steps: generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed; determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, and determining the second point cloud model corresponding to the point cloud relationship meeting a preset condition as a point cloud model matched with the first point cloud model, wherein the point cloud relationship comprises: scale factors, rotational matrices, and translations of transformations between the first point cloud model and the at least one second point cloud model; determining a key spraying path point and a spraying track of the first point cloud model based on a pre-configured key spraying path point of a matching point cloud model and the point cloud relation meeting a preset condition; and generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model.
In one embodiment, the point cloud relationship satisfying the preset condition includes at least one of: a scaling factor is closest to 1 in all point cloud relationships between the first point cloud model and the at least one second point cloud model; the rotation matrix is minimum in all point cloud relations between the first point cloud model and the at least one second point cloud model; the translation is minimal in all point cloud relationships between the first point cloud model and the at least one second point cloud model.
In one embodiment, the determining the key spraying path point and the spraying track of the first point cloud model based on the pre-configured key spraying path point of the matching point cloud model and the point cloud relationship satisfying the preset condition includes: extracting pre-configured key spraying path points of the matching point cloud model under a model coordinate system; converting the key spraying path points under the model coordinate system and configured key spraying path points into key spraying path points of the first point cloud model under a world coordinate system according to a scale factor, a rotation matrix and translation included in the point cloud relation meeting the preset condition; and connecting the key spraying path points of the first point cloud model to form a spraying track of the first point cloud model.
In one embodiment, the generating a robot control command for spraying the object to be sprayed according to the spraying trajectory of the first point cloud model includes: interpolating the spraying track of the first point cloud model at fixed time intervals to obtain robot joint set points; the robot joint set points are converted into robot control commands.
In one embodiment, the determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool comprises: determining a scale factor, a rotation matrix and a translation that minimize a matching error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model.
In one embodiment, the determining as the point cloud relationship between the first point cloud model and the at least one second point cloud model the scale factor, the rotation matrix, and the translation that will minimize the match error between the first point cloud model and the at least one second point cloud model comprises: detecting original key points of the first point cloud model and original key points of the at least one second point cloud model; filtering the original key points of the first point cloud model and the original key points of the at least one second point cloud model according to the quality of the original key points of the first point cloud model and the original key points of the at least one second point cloud model to obtain matchable key points of the first point cloud model and matchable key points of the at least one second point cloud model; extracting the characteristics of the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model; matching the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model according to the extracted characteristics of the matchable key points; and determining a point cloud relationship between the first point cloud model and the at least one second point cloud model according to the matching result.
In one embodiment, the quality of the original keypoint depends on its maximum curvature.
In one embodiment, the determining the point cloud relationship between the first point cloud model and the at least one second point cloud model according to the matching result comprises: estimating an initial point cloud relationship of the first point cloud model and the at least one second point cloud model using matching keypoints of the first point cloud model and the at least one second point cloud model, comprising: an initial scale factor, an initial rotation matrix and an initial translation as initial values for minimizing a matching error between the first point cloud model and the at least one second point cloud model by an iterative operation, wherein the iterative operation includes normal direction factors of matching key points of the first point cloud model and the at least one point cloud model; updating the point cloud relationship of the two points to carry out the iterative operation, and stopping the iterative operation until the iterative result meets the preset condition; and taking the point cloud relationship when the iterative operation is stopped as the point cloud relationship between the first point cloud model and the at least one second point cloud model.
In one embodiment, the predetermined condition satisfied by the iteration result includes: the ratio change rate of the scale factors between two adjacent iteration results, the angle difference change rate of the quaternion corresponding to the rotation matrix and the translation difference change rate are lower than the corresponding threshold values respectively.
The embodiment of the invention also provides an automatic spraying device, which comprises: the point cloud model generating module is used for generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed; the matching module is used for determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, and determining the second point cloud model corresponding to the point cloud relationship meeting a preset condition as the point cloud model matched with the first point cloud model, wherein the point cloud relationship comprises: scale factors, rotational matrices, and translations of transformations between the first point cloud model and the at least one second point cloud model; the spraying track determining module is used for determining key spraying path points and spraying tracks of the first point cloud model based on pre-configured key spraying path points of the matched point cloud model and the point cloud relation meeting the preset conditions; and the control command generation module is used for generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model.
The embodiment of the invention also provides an automatic spraying system which comprises the automatic spraying device.
Furthermore, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon; the computer program can be executed by a processor and implements the automatic spraying method as in the above embodiments.
According to the scheme of the invention, by constructing the model pool, the point cloud model of the object to be sprayed can be input, and the matching point cloud model matched with the point cloud model can be found from the model pool. And then obtaining key spraying path points in the point cloud model of the object to be sprayed according to the pre-configured key spraying path points of the matched point cloud model and the point cloud relation or transformation relation between the two point cloud models, and generating a spraying track and a robot control command according to the key spraying path points so as to spray the object to be sprayed. Therefore, for example, when a new product is sprayed, the spraying track can be generated without additional programming on the robot, and the robot control command can be automatically generated, so that the spraying and production cost is reduced, and the spraying and production efficiency is improved.
Drawings
The foregoing and other features and advantages of the invention will become more apparent to those skilled in the art to which the invention relates upon consideration of the following detailed description of a preferred embodiment of the invention with reference to the accompanying drawings, in which:
fig. 1 is a schematic flow chart of an automatic spraying method in an embodiment of the present invention.
Fig. 2 is a schematic diagram of a point cloud model of a chair, and key spray path points and spray tracks according to an embodiment of the invention.
FIG. 3 is a schematic flow chart of point cloud model matching in the embodiment of the present invention.
Fig. 4 is a schematic view of an automatic spraying device according to an embodiment of the present invention.
Fig. 5 is a schematic view of another automatic coating device according to an embodiment of the present invention.
Fig. 6 is a schematic view of an automatic spray system in an embodiment of the invention.
Fig. 7 is a schematic diagram of a specific application scenario of the scheme in the embodiment of the present invention.
Detailed Description
Fig. 1 is a flowchart illustrating an example of an automatic spraying method according to an embodiment of the present invention. The method may be used with a variety of computer devices. As shown in fig. 1, the method may include the steps of:
step S101, generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed.
Wherein, the object to be sprayed is furniture, product fittings and the like. Such as spray paint, etc. The point cloud information of the object to be painted is captured by, for example, a 3D (three-dimensional) sensor, may be captured in a dynamic process, for example, during conveyance of the object to be painted by a conveyor belt, or may be captured in a static process, for example, during placement of the object to be painted on a table or on the ground.
The point cloud information captured by the 3D sensor includes point cloud information of the object to be painted and may also include point cloud information of a background in which the object to be painted is located. For example, when spraying furniture, the furniture is usually placed on the ground or on a conveyor belt, and when obtaining the point cloud information of an object to be sprayed, the point cloud information of the ground or the point cloud information of the conveyor belt is obtained at the same time. In this case, point cloud information of the background may be identified based on the background pattern and removed to extract point cloud information of the object to be sprayed from the point cloud information captured by the 3D sensor. And then, removing noise point cloud data in the point cloud information of the object to be sprayed through a clustering algorithm. After the point cloud information of the object to be sprayed is obtained, a first point cloud model of the object to be sprayed can be generated according to the point cloud information of the object to be sprayed.
Step S102, determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, and determining the second point cloud model corresponding to the point cloud relationship meeting preset conditions as a point cloud model matched with the first point cloud model, wherein the point cloud relationship comprises: scale factors, rotational matrices, and translations of the transformation between the first point cloud model and the at least one second point cloud model.
According to an embodiment of the present invention, the pre-constructed model pool may include at least one second point cloud model, which may be generated, for example, in a manner similar to the generation of the first point cloud model in step S101. After the model pool is constructed, an experienced spray worker may enter the critical spray path points for each second point cloud model as pre-configured critical spray path points, stored with each second point cloud model, e.g., on the computer device, or on a storage device accessible by the computer device. These preconfigured key spray path points are used to indicate the points on the spray trajectory and the order of passage of each second point cloud model. Fig. 2 is a point cloud model of a chair, along with key spray path points and spray trajectories, in an embodiment of the invention. In fig. 2, the point cloud model of the chair includes a large number of points, forming the shape of the chair, circles represent key spray path points, and arrows represent spray trajectories and the order of passing the key spray path points.
Searching the model pool for a model matching the first point cloud model given the first point cloud model. During the searching, a point cloud relationship between the first point cloud model and each second point cloud model in the model pool may be determined, the point cloud relationship comprising: scale factors, rotational matrices, and translations of the transformation between the first point cloud model and the at least one second point cloud model.
Determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool may include: determining a scale factor, a rotation matrix and a translation that minimize a matching error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model.
Specifically, for example, given a first point cloud model PtCloud1 and a second point cloud model PtCloud2, by minimizing the error between the two point cloud models, the corresponding scale factor, rotation matrix, and translation are obtained. The process can be described as:
wherein e (s, R, T) is the error between PtCloud1 and PtCloud2, s is the scale factor between the two, R is the rotation matrix between the two, and T is the translation vector between the two; kp1jThe matching key point in PtCloud1, kp2jIs the matching key point in PtCloud2, n is the number of matching key points, "| | |" indicates the squaring operation, and the following is similar. The matching key is the point selected from the PtCloud1 and PtCloud2 at which two model matches are made.
And determining a second point cloud model corresponding to the point cloud relation meeting the preset condition as the point cloud model matched with the first point cloud model. Wherein the preset condition comprises at least one of: a scaling factor is closest to 1 in all point cloud relationships between the first point cloud model and the at least one second point cloud model; the rotation matrix is minimum in all point cloud relations between the first point cloud model and the at least one second point cloud model; the translation vector is smallest in all point cloud relations between the first point cloud model and the at least one second point cloud model.
Step S103, determining key spraying path points and spraying tracks of the first point cloud model based on pre-configured key spraying path points of the matched point cloud model and the point cloud relation meeting the preset conditions.
The step may further include: extracting pre-configured key spraying path points of the matching point cloud model under a model coordinate system; converting the key spraying path points under the model coordinate system and configured key spraying path points into key spraying path points of the first point cloud model under a world coordinate system according to a scale factor, a rotation matrix and translation included in the point cloud relation meeting the preset condition; and connecting the key spraying path points of the first point cloud model to form a spraying track of the first point cloud model. In addition, a blending operation can be performed at the position where the corner appears in the spraying track, so that the robot can smoothly spray at the corner, and the spraying is more uniform.
As previously described, its preconfigured critical spray path is stored with each second point cloud model. Therefore, after the second point cloud model matching the first point cloud model is determined, the preconfigured key spraying path points of the matching second point cloud model can be extracted, for example, the karman coordinates of the preconfigured key spraying path points in the model coordinate system are obtained. Then, the key spraying path point of the first point cloud model can be calculated according to the karman coordinate of the preconfigured key spraying path point in the model coordinate system and the scale factor, the rotation matrix and the translation vector (or the translation matrix) when the first point cloud model and the matching point cloud model are transformed, for example, the coordinate of the key spraying path point of the first point cloud model in the world coordinate system is calculated. Then, a spraying trajectory of the first point cloud model, which passes through all key spraying path points of the first point cloud model in sequence, may be calculated using, for example, a spline curve (spline curve) by considering a spraying speed parameter or the like input by the user.
And step S104, generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model.
This step may include: interpolating the spraying track of the first point cloud model at fixed time intervals to obtain robot joint set points; the robot joint set points are converted into robot control commands. Wherein the fixed time interval is, for example, 8ms (milliseconds). The robot joint set points include, for example: robot joint position, velocity, etc. Through converting robot joint setpoint into robot control command to after giving the robot, can control the removal of robot, carry out the spraying.
By the automatic spraying method, the experience of workers can be digitalized through the experience model pool, the spraying track can be generated without additional programming on the robot, the robot control command can be automatically generated, the spraying and production cost is reduced, and the spraying and production efficiency is improved. The method of the embodiment of the invention can be applied to furniture with various complex shapes and the like.
Fig. 3 is a flowchart illustrating an exemplary point cloud model matching process in an embodiment of the present invention. Referring to fig. 3, the determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool in step S102, and determining a second point cloud model corresponding to the point cloud relationship satisfying a preset condition as a point cloud model matching the first point cloud model may include:
step S301, detecting an original key point of the first point cloud model and an original key point of the at least one second point cloud model.
The key points are high-robustness 3D points with local invariance and noise resistance on a 3D point cloud or a model. The key points are not representational or descriptive, and the key points are used instead of all the points on the original point cloud, so that the data processing amount can be reduced, and the data processing speed can be increased. The detection of the original key points of the first point cloud model PtCloud1 is described below as an example, and the detection of the original key points of the second point cloud model PtCloud2 is similar.
Using sample point v in the first point cloud model PtCloud1 as the center of sphere, with different radii rkA series of local surface vectors Vector < LS1 are intercepted or created from a first point cloud model PtCloudlk＞。LS1kIs that the radius is rkIs a 3 XN set of pointskWherein N iskK may be a positive integer, being the number of points on the sphere.
Based on the local surface vector LS1kUsing Hotelling transform, a local coordinate system with rigid transformation invariance is established. Among them, a transform method of transforming a set of discrete signals into an uncorrelated sequence is called a Hotelling transform. When the object model performs rotation or translation, the local coordinate system also performs corresponding rotation or translation operations. LS1 may be calculated using Principal Component Analysis (PCA) prior to performing the Hotelling transformkThe feature vector of (2). The feature vector may also be calculated in other ways. In calculating the feature vector, LS1 is calculated firstkAnd calculating the covariance matrix, the mean value needs to be calculated. The calculation process for the feature vector is described as follows:
LS1kaverage value of points in (1)
Comprises the following steps:
wherein, LS1kiIs LS1kPoint (x) of (1)i，yi，zi)T，i＝1，2......，Nk，NkIs that the radius is rkThe number of points on the sphere. LS1kCovariance matrix C ofkComprises the following steps:
for covariance matrix CkUsing principal component analysis, CkThe eigenvalue (eigenvalue) of (c) is decomposed into:
Ck*Vk＝Dk*Vk (4)
wherein, VkIs a feature vector matrix, DkIs CkA diagonal matrix of eigenvalues of (a).
For vk＝[vk1，vk2，vk3](vk1、vk2、vk3Is VkThe feature vector on the diagonal line) of the first point cloud model, the curvature of which corresponding to the sample point v on the first point cloud model is defined as
Normalized vkIs a normal to the point v
Solving for local surface vector LS1kAfter the feature vector of (3), the matrix LS1 can be transformed based on Hotelling, using equation (5)kAligned with its principal axis to obtain a transformed local surface vector
Wherein the content of the first and second substances,
representing transformed local surface vectors
The local surface vector of the ith point is from local surface vector LS1kThe ith point LS1kiTransformed into.
By varying the radius r of the spherekVector < LS1 can be obtainedkMultiple LS1 ink. The best LS1 is selected by the following rulek。
The positive direction of the x axis is specified as the point cloud dense direction. The metric value alphakIs defined as equation (6) and is referred to as the ratio between the first two principal axes, which are the x-axis and the y-axis of the transformed local coordinate system.
Wherein the content of the first and second substances,
and
representing transformed local surface vectors
The x-coordinate and y-coordinate set of the point in (1).
For Vector < LS1kMaximum value ofkAnd recording the corresponding sample point v as a key point (namely obtaining the original key point).
Step S302, filtering the original key points of the first point cloud model and the original key points of the at least one second point cloud model according to the quality of the original key points of the first point cloud model and the original key points of the at least one second point cloud model to obtain matchable key points of the first point cloud model and matchable key points of the at least one second point cloud model.
According to an embodiment of the invention, after obtaining the original keypoints of said first point cloud model PtCloud1 and of said at least one second point cloud model PtCloud2, the quality thereof (distance between the original keypoints) is calculated by calculating the curvature of said original keypoints, said quality of the original keypoints depending for example on their maximum curvature, as follows:
wherein Q isiIs the quality of the ith original keypoint, ciIs the curvature of the ith original keypoint, w1、w2For weighting, N is the number of original keypoints. By setting the quality threshold _ Q, Q can be comparediWith the quality threshold _ q, the original keypoints with quality greater than the quality threshold are filtered out and retained as matchable keypoints.
Step S303, extracting the characteristics of the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model.
According to the embodiment of the application, the curved surface pair can be used
The points in (around the keypoint v) are fitted to form a 3D surface. The fitted 3D surface may be sampled using an n × n grid during fitting. For each matchable keypoint, extracting n2The z-axis value of each grid point is used as its depth feature _ depth. The feature depth is 1 xn2The vector of (2) is described to be unique. All feature vectors of 1 xn 2 for all matchable keypoints are normalized to ensure that they are not affected by the scale factor. Reducing the 1 xn using PCA2The vector of (c) finally obtains matchable for PtCloud1, PtCloud2And matching a 1 multiplied by L vector of the characteristics of the key points, wherein L is a positive integer.
Step S304, matching the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model according to the extracted features of the matchable key points to obtain the matched key points of the first point cloud model and the second point cloud model.
According to the embodiment of the present invention, after extracting the features of the matchable keypoints of the first point cloud model PtCloud1 and the at least one second point cloud model PtCloud2, each feature of PtCloud1 and PtCloud2 is matched by minimizing the error metric:
ferrmin＝min(ferrij) (8)
wherein, ferrijRepresents the error metric, fer, between the feature of the ith matchable keypoint in PtCloud1 and the feature of the jth matchable keypoint in PtCloud2minRepresenting a measure of error ferrijI, j are positive integers.
The error metric between the two features is as follows:
ferrij＝cos-1(|fpc1i*fpc2j|) (9)
wherein f ispc1iFeature representing the ith matchable keypoint in PtCloud1, fpc2jFeatures that represent the jth matchable keypoint in PtCloud 2. The matching key points matched between the first point cloud model and the second point cloud model can be obtained through the method. For example, by ferrijCalculating error metrics between the features of the ith matchable keypoint in PtCloud1 and the features of each matchable keypoint in PtCloud2, and determining an error metric minimum ferminThe characteristics of the matchable keypoints in the corresponding PtCloud2, thereby determining the matchable keypoints in PtCloud2 corresponding to the ith matchable keypoint in PtCloud 1.
Step S305, determining a point cloud relationship between the first point cloud model and the at least one second point cloud model according to a matching result.
Step S305 may specifically include: estimating an initial point cloud relationship of the first point cloud model and the at least one second point cloud model using matching keypoints of the first point cloud model and the at least one second point cloud model, comprising: an initial scale factor, an initial rotation matrix and an initial translation as initial values for minimizing a matching error between the first point cloud model and the at least one second point cloud model by an iterative operation, wherein the iterative operation includes normal direction factors of matching key points of the first point cloud model and the at least one point cloud model; updating the point cloud relationship of the two points to carry out the iterative operation, and stopping the iterative operation until the iterative result meets the preset condition; and taking the point cloud relationship when the iterative operation is stopped as the point cloud relationship between the first point cloud model and the at least one second point cloud model. Wherein the predetermined condition satisfied by the iteration result includes: the ratio change rate of the scale factors between two adjacent iteration results, the angle difference change rate of the quaternion corresponding to the rotation matrix and the translation difference change rate are lower than the corresponding threshold values respectively.
According to an embodiment of the invention, the point cloud relationship between the first point cloud model PtCloud1 and the at least one second point cloud model PtCloud2 is determined by iteratively minimizing equation (1) by changing s, R, T in equation (1).
An initial iteration value is first determined and then a final value is found by iteration. The process is described as follows:
(1) determining initial iteration values
Estimating initial transformation parameters using matching keypoints of the first point cloud model PtCloud1 and the at least one second point cloud model PtCloud2, comprising: initial scale factor s0Initial rotation matrix R0And an initial translation matrix T0。
Centroids are calculated for two sets of matching keypoints (e.g., kp1, kp2, respectively) for PtCloud1 and PtCloud2, transforming the matching keypoints into transformed keypoints centered on the mean. The transformation is performed by:
wherein kp1jAnd kp2jRepresents the matching key points in PtCloud1 and PtCloud 2;
is the centroid of kp1, kp 2;
and
are transformed keypoints centered on the mean. Calculating distances of all transformed keypoints of PtCloud1, PtCloud2, respectively (e.g., each transformed keypoint
Distances to the origin of the coordinate system), for example, sum of sum _ kp1, sum _ kp2, and their ratio sum _ kp2/sum _ kp1 is calculated as the initial scale factor s0。
Then, the initial rotation matrix R is calculated0. Ideally, there is no error between the two point cloud models after registration, i.e., e (s, R, T) ═ 0 (0 on the right side of the equation (1)). Thus, it is possible to obtain:
where T represents a translation matrix.
Substituting equation (11) into equation (1) results in equation (12):
where e (s, R) represents the error between PtCloud1 and PtCloud2, and n is the total number of matching keypoints.
Here, the number of the first and second electrodes,s0for a better initial scaling factor, which can minimize e (s, R), it has been calculated previously. In addition, equation (12) can be split into two parts: some are terms containing R and some are not. The split error e (s, R) can be rewritten as:
singular Value Decomposition (SVD) methods can be used such that
Maximize and minimize equation (13), the resulting R value is the initial rotation matrix R0。
Finally, s is0And R0Input to equation (11) in place of S and R is used to calculate the initial translation vector T0.
(2) Finding the final value by iteration
Given an initial iteration value(s)0、R0And T0) And executing a search process to find a final S, R, T value through iteration, wherein the specific process is as follows:
s, R, T initialized to s0、R0And T0. Then, s is updated in an iteration (i ═ 0, 1.., cnt-1) according to equation (14) belowi、RiAnd TiAnd cnt is the number of iterations.
The iterative process is as follows:
for i＝1；i＜cnt；i++
si＝si-1+Δs
Ri＝Ri-1+ΔR
Ti＝Ti-1+ΔT
calculation of e (s, R, T)
Wherein the content of the first and second substances,Δ s, Δ R, Δ T are each an iteration increment of S, R, T.
Is the normal to the matching keypoints in PtCloud1 and PtCloud 2.
See above for calculations
Kp1 hereinj、kp2jIs an embodiment of v.
Equation (14) considers the normal directions of the matching keypoints of PtCloud1 and PtCloud2 on the basis of equation (1). Specifically, equation (14) includes pairing kp1 by rotating matrix RjNormal to
Rotating the cylinder to obtain the rotation number of the cylinder and the kp2jNormal to
The second-order mode of the difference of (a), i.e. the sum of the squares of the differences of the three components (x, y, z components) of the two normal vectors.
The stopping rule for the iteration is as follows:
1): will rotate the matrix RiConversion to quaternion quati；
2): according to equation (15), a quaternion quat of the results of two adjacent iterations (e.g., the ith iteration result and the (i + 1) th iteration result) is calculatedi、quati+1Angle difference between delta thetai+1And rate of change of angular difference
Where "dot" is the dot product between vectors.
3): calculating the translation T between two adjacent iteration resultsi+1、TiDifference value Δ T ofi+1And rate of change of translational difference
4): calculating a scaling factor s between two adjacent iteration resultsi+1、siIs a ratio ofi+1And rate of change of ratio
5): when all the change rates
And stopping iteration when the values are all smaller than the threshold values. I.e. the rate of change of the ratio of the scaling factors between two adjacent iteration results
Rate of change of angular difference of quaternion corresponding to rotation matrix
And rate of change of translational difference value
And stopping iteration when the values are lower than the corresponding threshold values. If the iteration count cnt is reached without satisfying the iteration stop condition, the result with the smallest e (s, R, T) in all iterations can be selected, or the iteration increment is appropriately added and recalculated. Taking the final S, R, T value when the iteration stops as the first point cloudA point cloud relationship, i.e., a transformation relationship, between the model and the second point cloud model.
And then, repeating the above processes to obtain point cloud relations between the first point cloud model and all the second point cloud models in the model pool, selecting the point cloud relations meeting preset conditions, and taking the corresponding second point cloud models as matching models of the first point cloud model.
Correspondingly, an embodiment of the present invention further provides an automatic spraying device, as shown in fig. 4, the automatic spraying device may include: a point cloud model generation module 401, a matching module 402, a spraying track determination module 403 and a control command production module 404.
The point cloud model generating module 401 is configured to generate a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed.
A matching module 402, configured to determine a point cloud relationship between the first point cloud model and at least one second point cloud model in a preconfigured model pool, and determine a second point cloud model corresponding to the point cloud relationship meeting a preset condition as a point cloud model matched with the first point cloud model, where the point cloud relationship includes: scale factors, rotational matrices, and translations of the transformation between the first point cloud model and the at least one second point cloud model. According to an embodiment of the present application, for example, the point cloud relationship satisfying the preset condition includes at least one of: a scaling factor is closest to 1 in all point cloud relationships between the first point cloud model and the at least one second point cloud model; the rotation matrix is minimum in all point cloud relations between the first point cloud model and the at least one second point cloud model; the translation is minimal in all point cloud relationships between the first point cloud model and the at least one second point cloud model. The matching module 402 determines a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, including: determining a scale factor, a rotation matrix and a translation that minimize a matching error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model.
More specifically, the matching module 402 determines a scale factor, a rotation matrix, and a translation that minimize a matching error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model comprises: detecting original key points of the first point cloud model and original key points of the at least one second point cloud model; filtering the original key points of the first point cloud model and the original key points of the at least one second point cloud model according to the quality of the original key points of the first point cloud model and the original key points of the at least one second point cloud model to obtain matchable key points of the first point cloud model and matchable key points of the at least one second point cloud model; extracting the characteristics of the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model; matching the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model according to the extracted characteristics of the matchable key points; and determining a point cloud relationship between the first point cloud model and the at least one second point cloud model according to the matching result. Wherein the quality of the original keypoint depends, for example, on its maximum curvature.
According to an embodiment of the present application, the determining, by the matching module 402, a point cloud relationship between the first point cloud model and the at least one second point cloud model according to the matching result includes: estimating an initial point cloud relationship of the first point cloud model and the at least one second point cloud model using matching keypoints of the first point cloud model and the at least one second point cloud model, comprising: an initial scale factor, an initial rotation matrix and an initial translation as initial values for minimizing a matching error between the first point cloud model and the at least one second point cloud model by an iterative operation, wherein the iterative operation includes normal direction factors of matching key points of the first point cloud model and the at least one point cloud model; updating the point cloud relationship of the two points to carry out the iterative operation, and stopping the iterative operation until the iterative result meets the preset condition; and taking the point cloud relationship when the iterative operation is stopped as the point cloud relationship between the first point cloud model and the at least one second point cloud model. The predetermined conditions satisfied by the iteration result include, for example: the ratio change rate of the scale factors between two adjacent iteration results, the angle difference change rate of the quaternion corresponding to the rotation matrix and the translation difference change rate are lower than the corresponding threshold values respectively.
A spraying track determining module 403, configured to determine a key spraying path point and a spraying track of the first point cloud model based on a pre-configured key spraying path point of the matching point cloud model and the point cloud relationship meeting the preset condition. Wherein, in particular, the spraying trajectory determination module 403 may be configured to extract preconfigured key spraying path points of the matching point cloud model under a model coordinate system; converting the key spraying path points under the model coordinate system and configured key spraying path points into key spraying path points of the first point cloud model under a world coordinate system according to a scale factor, a rotation matrix and translation included in the point cloud relation meeting the preset condition; and connecting the key spraying path points of the first point cloud model to form a spraying track of the first point cloud model.
And a control command generating module 404, configured to generate a robot control command for spraying the object to be sprayed according to the spraying trajectory of the first point cloud model. Specifically, the control command generating module 404 may be configured to interpolate the spraying trajectory of the first point cloud model at fixed time intervals, so as to obtain a robot joint set point; the robot joint set points are converted into robot control commands.
By the automatic spraying device provided by the embodiment of the invention, the experience of a worker can be digitalized through the experience model pool, the spraying track can be generated without additionally programming the robot, the robot control command can be automatically generated, the spraying and production cost is reduced, and the spraying and production efficiency is improved. The device of the embodiment of the invention can be applied to furniture and the like with various complex shapes.
Fig. 5 is a schematic view of another automatic coating device according to an embodiment of the present invention. As shown in fig. 5, may include: at least one memory 51 and at least one processor 52. In addition, some other components may be included, such as a communications port, etc. These components communicate over a bus.
Wherein the at least one memory 51 is adapted to store a computer program. In one embodiment, the computer program may be understood to include the various modules shown in FIG. 4. Further, the at least one memory 51 may also store an operating system and the like. Operating systems include, but are not limited to: an Android operating system, a Symbian operating system, a Windows operating system, a Linux operating system, and the like.
The at least one processor 52 is configured to invoke a computer program stored in the at least one memory 51 to perform the automated spray method described in the embodiments of the present invention. The processor 52 may be a CPU, processing unit/module, ASIC, logic module, or programmable gate array, etc. Which can receive and transmit data through the communication port.
The embodiment of the invention also provides an automatic spraying system. As shown in fig. 6, the automatic spray system includes: 3D sensor 601, automatic spray apparatus 602, spray robot 603, model pool storage 604.
Among other things, the 3D sensor 601 is used to capture point cloud information of the object 600 to be painted (e.g., various furniture), for example. The automated spray coating device 602 may include any of the automated spray coating devices described above. After generating the robot control command, the automatic spraying device issues the command to the spraying robot 603 to spray the object 600 to be sprayed. At least one second point cloud model is stored in the model pool storage 604 for matching with the first point cloud model of the object 600 to be painted. The painting robot 603 may be a painting robot system including a painting tool.
Fig. 7 is a schematic diagram of a specific application scenario of the scheme in the embodiment of the present invention. As shown in fig. 7, the application scenario is to paint furniture 701, such as a table, a chair, and the like. In this scenario, furniture 701 is placed on a conveyor belt 702, which is transported by the conveyor belt 702 to a painting robot 704 for painting. The conveyor belt 702 is driven by a motor or electric machine 703. The 3D sensor 705 may capture point cloud information for furniture. The system controller 706 may include the aforementioned automatic spraying apparatus of the present invention, receive the point cloud information of the furniture captured by the 3D sensor 705, and after processing, generate a robot control command and send the robot control command to the spraying robot 704. The user interface and parameter input may use a Human Machine Interface (HMI) 707. A programmable controller (PLC) or an industrial computer (IPC) is used as the system controller. System programming (including TIA robot library for controlling robots) is performed using fully integrated automated routing software (TIA Portal). In this application scenario, the exchange of information, the robot control command, and the update of status may use Profinet protocol. For example, via Profinet protocol, the system controller 706 controls the motor 703 to move the conveyor belt, information exchange between the 3D sensor 705 and the system controller 706, information exchange between the system controller 706 and the spray robot 704, and the like. The Profinet protocol is an automation bus standard based on industrial ethernet technology.
It should be noted that not all steps and modules in the above flows and structures are necessary, and some steps or modules may be omitted according to actual needs. The execution order of the steps is not fixed and can be adjusted as required. The division of each module is only for convenience of describing adopted functional division, and in actual implementation, one module may be divided into multiple modules, and the functions of multiple modules may also be implemented by the same module, and these modules may be located in the same device or in different devices.
It is understood that the hardware modules in the above embodiments may be implemented mechanically or electronically. For example, a hardware module may include a specially designed permanent circuit or logic device (e.g., a special purpose processor such as an FPGA or ASIC) for performing specific operations. A hardware module may also include programmable logic devices or circuits (e.g., including a general-purpose processor or other programmable processor) that are temporarily configured by software to perform certain operations. The implementation of the hardware module in a mechanical manner, or in a dedicated permanent circuit, or in a temporarily configured circuit (e.g., configured by software), may be determined based on cost and time considerations.
In addition, a computer-readable storage medium is provided in the embodiments of the present invention, and has a computer program stored thereon, where the computer program can be executed by a processor and implement the automatic spraying method described in the embodiments of the present invention. Specifically, a system or an apparatus equipped with a storage medium on which a software program code that realizes the functions of any of the embodiments described above is stored may be provided, and a computer (or a CPU or MPU) of the system or the apparatus is caused to read out and execute the program code stored in the storage medium. Further, part or all of the actual operations may be performed by an operating system or the like operating on the computer by instructions based on the program code. The functions of any of the above-described embodiments may also be implemented by writing the program code read out from the storage medium to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion unit connected to the computer, and then causing a CPU or the like mounted on the expansion board or the expansion unit to perform part or all of the actual operations based on the instructions of the program code. Examples of the storage medium for supplying the program code include floppy disks, hard disks, magneto-optical disks, optical disks (e.g., CD-ROMs, CD-R, CD-RWs, DVD-ROMs, DVD-RAMs, DVD-RWs, DVD + RWs), magnetic tapes, nonvolatile memory cards, and ROMs. Alternatively, the program code may be downloaded from a server computer via a communications network.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (12)
1. An automatic spray coating method, comprising:
generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed;
determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, and determining the second point cloud model corresponding to the point cloud relationship meeting a preset condition as a point cloud model matched with the first point cloud model, wherein the point cloud relationship comprises: scale factors, rotational matrices, and translations of transformations between the first point cloud model and the at least one second point cloud model;
determining a key spraying path point and a spraying track of the first point cloud model based on a pre-configured key spraying path point of a matching point cloud model and the point cloud relation meeting a preset condition;
and generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model.
2. The method according to claim 1, wherein the point cloud relationship satisfying the preset condition comprises at least one of:
a scaling factor is closest to 1 in all point cloud relationships between the first point cloud model and the at least one second point cloud model;
the rotation matrix is minimum in all point cloud relations between the first point cloud model and the at least one second point cloud model;
the translation is minimal in all point cloud relationships between the first point cloud model and the at least one second point cloud model.
3. The method of claim 1 or 2, wherein determining the critical spray path points and spray trajectory of the first point cloud model based on the pre-configured critical spray path points of the matching point cloud model and the point cloud relationship satisfying the preset condition comprises:
extracting pre-configured key spraying path points of the matching point cloud model under a model coordinate system;
converting the key spraying path points under the model coordinate system and configured key spraying path points into key spraying path points of the first point cloud model under a world coordinate system according to a scale factor, a rotation matrix and translation included in the point cloud relation meeting the preset condition;
and connecting the key spraying path points of the first point cloud model to form a spraying track of the first point cloud model.
4. The method according to claim 3, wherein the generating of the robot control command for spraying the object to be sprayed according to the spraying trajectory of the first point cloud model comprises:
interpolating the spraying track of the first point cloud model at fixed time intervals to obtain robot joint set points;
the robot joint set points are converted into robot control commands.
5. The method of claim 1, wherein determining the point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool comprises:
determining a scale factor, a rotation matrix and a translation that minimize a matching error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model.
6. The method of claim 5, wherein the determining a scale factor, a rotation matrix, and a translation that will minimize a match error between the first point cloud model and the at least one second point cloud model as a point cloud relationship between the first point cloud model and the at least one second point cloud model comprises:
detecting original key points of the first point cloud model and original key points of the at least one second point cloud model;
filtering the original key points of the first point cloud model and the original key points of the at least one second point cloud model according to the quality of the original key points of the first point cloud model and the original key points of the at least one second point cloud model to obtain matchable key points of the first point cloud model and matchable key points of the at least one second point cloud model;
extracting the characteristics of the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model;
matching the matchable key points of the first point cloud model and the matchable key points of the at least one second point cloud model according to the extracted characteristics of the matchable key points; and
and determining the point cloud relation between the first point cloud model and the at least one second point cloud model according to the matching result.
7. The method of claim 6, wherein the quality of the original keypoint depends on its maximum curvature.
8. The method according to claim 6 or 7, wherein the determining a point cloud relationship between the first point cloud model and the at least one second point cloud model according to the matching result comprises:
estimating an initial point cloud relationship of the first point cloud model and the at least one second point cloud model using matching keypoints of the first point cloud model and the at least one second point cloud model, comprising: an initial scale factor, an initial rotation matrix and an initial translation as initial values for minimizing a matching error between the first point cloud model and the at least one second point cloud model by an iterative operation, wherein the iterative operation includes normal direction factors of matching key points of the first point cloud model and the at least one point cloud model;
updating the point cloud relationship of the two points to carry out the iterative operation, and stopping the iterative operation until the iterative result meets the preset condition;
and taking the point cloud relationship when the iterative operation is stopped as the point cloud relationship between the first point cloud model and the at least one second point cloud model.
9. The method of claim 8, wherein the predetermined condition satisfied by the iteration result comprises:
the ratio change rate of the scale factors between two adjacent iteration results, the angle difference change rate of the quaternion corresponding to the rotation matrix and the translation difference change rate are lower than the corresponding threshold values respectively.
10. An automatic spray coating device, comprising:
the point cloud model generating module is used for generating a first point cloud model of the object to be sprayed according to the point cloud information of the object to be sprayed;
the matching module is used for determining a point cloud relationship between the first point cloud model and at least one second point cloud model in a pre-constructed model pool, and determining the second point cloud model corresponding to the point cloud relationship meeting a preset condition as the point cloud model matched with the first point cloud model, wherein the point cloud relationship comprises: scale factors, rotational matrices, and translations of transformations between the first point cloud model and the at least one second point cloud model;
the spraying track determining module is used for determining key spraying path points and spraying tracks of the first point cloud model based on pre-configured key spraying path points of the matched point cloud model and the point cloud relation meeting the preset conditions;
and the control command generation module is used for generating a robot control command for spraying the object to be sprayed according to the spraying track of the first point cloud model.
11. An automatic coating system comprising the automatic coating device according to claim 10.
12. A computer-readable storage medium having stored thereon a computer program; characterized in that the computer program is executable by a processor and implements the automatic spraying method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011037006.8A CN114274139B (en) | 2020-09-27 | 2020-09-27 | Automatic spraying method, device, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011037006.8A CN114274139B (en) | 2020-09-27 | 2020-09-27 | Automatic spraying method, device, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114274139A true CN114274139A (en) | 2022-04-05 |
| CN114274139B CN114274139B (en) | 2024-04-19 |
Family
ID=80868014
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011037006.8A Active CN114274139B (en) | 2020-09-27 | 2020-09-27 | Automatic spraying method, device, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114274139B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106023156A (en) * | 2016-05-10 | 2016-10-12 | 中国商用飞机有限责任公司 | Point-cloud model and CAD model registering method based on detection features |
| CN106570903A (en) * | 2016-10-13 | 2017-04-19 | 华南理工大学 | Visual identification and positioning method based on RGB-D camera |
| GB201714179D0 (en) * | 2017-09-05 | 2017-10-18 | Nokia Technologies Oy | Cross-source point cloud registration |
| CN108038908A (en) * | 2017-11-21 | 2018-05-15 | 泰瑞数创科技(北京)有限公司 | Spatial object identification and modeling method and system based on artificial intelligence |
| CN108052624A (en) * | 2017-12-15 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Processing Method of Point-clouds, device and computer readable storage medium |
| DE102016123149A1 (en) * | 2016-11-30 | 2018-05-30 | Dornheim Medical Images GmbH | IMAGE-DATA-BASED RECONSTRUCTION OF THREE-DIMENSIONAL SURFACES |
| CN108161941A (en) * | 2018-03-06 | 2018-06-15 | 广东工业大学 | A kind of robot trajectory's optimization method and device for applying to timber spraying industry |
| CN109454642A (en) * | 2018-12-27 | 2019-03-12 | 南京埃克里得视觉技术有限公司 | Robot coating track automatic manufacturing method based on 3D vision |
| CN109766903A (en) * | 2018-08-13 | 2019-05-17 | 西安理工大学 | A kind of point cloud model SURFACES MATCHING method based on curved surface features |
| CN110039561A (en) * | 2019-05-14 | 2019-07-23 | 南京理工大学 | Hot line robot remote operating staff training system and method based on cloud |
| US20190255551A1 (en) * | 2018-02-22 | 2019-08-22 | Hope Robotics Llc | Autonomous mobile coating applicator |
| CN110288638A (en) * | 2019-06-18 | 2019-09-27 | 济南大学 | A kind of knochenbruch model rough registration method, system and knochenbruch Model registration method |
| CN111013883A (en) * | 2019-11-25 | 2020-04-17 | 浙江明泉工业涂装有限公司 | Robot control method for intelligent spraying of multiple vehicle types |
| CN111508008A (en) * | 2020-04-08 | 2020-08-07 | 达闼机器人有限公司 | Point cloud registration method, electronic equipment and storage medium |
| CN111652801A (en) * | 2020-05-11 | 2020-09-11 | 东莞理工学院 | Accurate point cloud splicing method |
-
2020
- 2020-09-27 CN CN202011037006.8A patent/CN114274139B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106023156A (en) * | 2016-05-10 | 2016-10-12 | 中国商用飞机有限责任公司 | Point-cloud model and CAD model registering method based on detection features |
| CN106570903A (en) * | 2016-10-13 | 2017-04-19 | 华南理工大学 | Visual identification and positioning method based on RGB-D camera |
| DE102016123149A1 (en) * | 2016-11-30 | 2018-05-30 | Dornheim Medical Images GmbH | IMAGE-DATA-BASED RECONSTRUCTION OF THREE-DIMENSIONAL SURFACES |
| GB201714179D0 (en) * | 2017-09-05 | 2017-10-18 | Nokia Technologies Oy | Cross-source point cloud registration |
| CN108038908A (en) * | 2017-11-21 | 2018-05-15 | 泰瑞数创科技(北京)有限公司 | Spatial object identification and modeling method and system based on artificial intelligence |
| CN108052624A (en) * | 2017-12-15 | 2018-05-18 | 深圳市易成自动驾驶技术有限公司 | Processing Method of Point-clouds, device and computer readable storage medium |
| US20190255551A1 (en) * | 2018-02-22 | 2019-08-22 | Hope Robotics Llc | Autonomous mobile coating applicator |
| CN108161941A (en) * | 2018-03-06 | 2018-06-15 | 广东工业大学 | A kind of robot trajectory's optimization method and device for applying to timber spraying industry |
| CN109766903A (en) * | 2018-08-13 | 2019-05-17 | 西安理工大学 | A kind of point cloud model SURFACES MATCHING method based on curved surface features |
| CN109454642A (en) * | 2018-12-27 | 2019-03-12 | 南京埃克里得视觉技术有限公司 | Robot coating track automatic manufacturing method based on 3D vision |
| CN110039561A (en) * | 2019-05-14 | 2019-07-23 | 南京理工大学 | Hot line robot remote operating staff training system and method based on cloud |
| CN110288638A (en) * | 2019-06-18 | 2019-09-27 | 济南大学 | A kind of knochenbruch model rough registration method, system and knochenbruch Model registration method |
| CN111013883A (en) * | 2019-11-25 | 2020-04-17 | 浙江明泉工业涂装有限公司 | Robot control method for intelligent spraying of multiple vehicle types |
| CN111508008A (en) * | 2020-04-08 | 2020-08-07 | 达闼机器人有限公司 | Point cloud registration method, electronic equipment and storage medium |
| CN111652801A (en) * | 2020-05-11 | 2020-09-11 | 东莞理工学院 | Accurate point cloud splicing method |
Non-Patent Citations (1)
| Title |
|---|
| 陈驰;杨必胜;田茂;李健平;邹响红;吴唯同;宋易恒;: "车载MMS激光点云与序列全景影像自动配准方法", 测绘学报, no. 02, pages 215 - 224 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114274139B (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2928648B1 (en) | Robot program generation for robotic processes | |
| CN106391366B (en) | A kind of big component automation paint system of civil aircraft | |
| Lee et al. | An integrated framework for cooperative aerial manipulators in unknown environments | |
| US10540779B2 (en) | Posture positioning system for machine and the method thereof | |
| CN113920061A (en) | Industrial robot operation method and device, electronic equipment and storage medium | |
| Siradjuddin et al. | Image Based Visual Servoing of a 7 DOF robot manipulator using a distributed fuzzy proportional controller | |
| Navarro-Alarcon et al. | Visually servoed deformation control by robot manipulators | |
| Lai et al. | Integration of visual information and robot offline programming system for improving automatic deburring process | |
| Amersdorfer et al. | Equidistant tool path and cartesian trajectory planning for robotic machining of curved freeform surfaces | |
| CN113305847B (en) | Building 3D printing mobile mechanical arm station planning method and system | |
| CN111709095B (en) | Method for constructing 6D virtual clamp for complex curved surface | |
| CN114274139A (en) | Automatic spraying method, device, system and storage medium | |
| Ðurović et al. | Visual servoing for low-cost SCARA robots using an RGB-D camera as the only sensor | |
| Chen et al. | Transformative industrial robot programming in surface manufacturing | |
| Bayro-Corrochano et al. | The motor extended Kalman filter: A geometric approach for rigid motion estimation | |
| Thotakuri et al. | Survey on robot vision: techniques, tools and methodologies | |
| CN113492071B (en) | Path planning method applied to window frame spraying | |
| Al-Junaid | ANN based robotic arm visual servoing nonlinear system | |
| Nemer et al. | Off-line nominal path generation of 6-DoF robotic manipulator for edge finishing and inspection processes | |
| Shi et al. | Contouring error vector and cross-coupled control of multi-axis servo system | |
| Ito et al. | Feature-based visual target following for a nonholonomic wheeled mobile robot with a single camera | |
| JP4282874B2 (en) | Generating control data for manipulators | |
| Raible et al. | Automatic Path Planning for Robotic Grinding and Polishing Tasks based on Point Cloud Slicing | |
| CN109129558A (en) | A method of control mechanical arm tail end executes space curve movement | |
| Lu et al. | Paint surface estimation and trajectory planning for automated painting systems |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |