CN118247473A - Bounding box extraction method, device, electronic equipment and readable storage medium - Google Patents
Bounding box extraction method, device, electronic equipment and readable storage medium Download PDFInfo
- Publication number
- CN118247473A CN118247473A CN202410642952.7A CN202410642952A CN118247473A CN 118247473 A CN118247473 A CN 118247473A CN 202410642952 A CN202410642952 A CN 202410642952A CN 118247473 A CN118247473 A CN 118247473A
- Authority
- CN
- China
- Prior art keywords
- bounding box
- minimum
- minimum bounding
- vertex
- frames
- 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
- 238000000605 extraction Methods 0.000 title claims abstract description 34
- 238000001514 detection method Methods 0.000 claims description 12
- 238000000034 method Methods 0.000 claims description 10
- 230000000007 visual effect Effects 0.000 claims description 9
- 238000004891 communication Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- DMSMPAJRVJJAGA-UHFFFAOYSA-N benzo[d]isothiazol-3-one Chemical compound C1=CC=C2C(=O)NSC2=C1 DMSMPAJRVJJAGA-UHFFFAOYSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a bounding box extraction method, a bounding box extraction device, electronic equipment and a readable storage medium. And then obtaining the top points of the surface point clouds of the cargoes in each image frame, and obtaining the minimum bounding box of each cargoes. And detecting the minimum bounding box with the corresponding relation in the two frames of image frames based on the vertexes and the minimum bounding box of the surface point cloud in the two frames of image frames, and finally determining the target bounding box from the minimum bounding box with the corresponding relation. According to the scheme, the minimum bounding box with the corresponding relation in the two continuous image frames is detected, so that the target bounding box is determined, and the determined target bounding box can be ensured to be correct.
Description
Technical Field
The invention relates to the technical field of cargo stacking, in particular to a bounding box extraction method, a bounding box extraction device, electronic equipment and a readable storage medium.
Background
In a mixed palletizing scenario, the transported goods, such as boxes, are placed on a conveyor belt, and after each movement of the conveyor belt a fixed distance, the photographing apparatus may perform one image photographing to acquire image information containing the boxes. And acquiring the point cloud of the box based on the acquired image, and acquiring the position of the minimum bounding box of the box in the image through box subsurface point cloud extraction and minimum bounding box calculation. The determination of the minimum bounding box may be provided to an automated device for automated operations, such as grasping boxes, transferring boxes, etc., for palletizing.
Since some boxes are completely within the camera field of view and some boxes are partially within the camera field of view at the time of photographing, the box corresponding to the detected minimum bounding box is not necessarily a real box. For example, as shown in fig. 1, where the grey rectangle characterizes the box, the bold line shows the smallest bounding box, and the outermost rectangular box represents the camera field of view. By true boxes are meant boxes that subsequently enable the automation device to perform the automation operation accurately. In order to determine which boxes corresponding to the minimum bounding boxes in the image frames are real, a camera visual field range detection method is adopted in the conventional common mode, namely, which minimum bounding boxes are all in the camera visual field range is judged, the boxes corresponding to the minimum bounding boxes in the camera visual field range are determined to be real boxes, and then automatic equipment is controlled to grasp the boxes for stacking treatment.
This approach employed in the prior art suffers from drawbacks, for example, if there is a box of great length and width that comes into view of the camera, then in two consecutive image frames, the box's corresponding minimum bounding box is not entirely within the camera's view, as shown, for example, in fig. 2 (a). If camera field of view detection is used, then the bins in both frames will be determined to be false, however in practice the minimum bounding boxes of both bins are correct. Furthermore, if one side of a box is truncated by the camera field of view, a truncated box may be formed, for example, as shown in fig. 2 (b). The smallest bounding box of the truncated bin is entirely within the camera field of view. If the camera field of view detection method is employed, the minimum bounding box will be determined to be true, however in practice the minimum bounding box is incorrect.
Disclosure of Invention
The invention aims at providing a bounding box extraction method, a bounding box extraction device, electronic equipment and a readable storage medium, which can accurately determine a correct minimum bounding box.
Embodiments of the invention may be implemented as follows:
In a first aspect, the present invention provides a bounding box extraction method, the method comprising:
acquiring two continuous frames of image frames acquired by camera equipment, wherein each image frame has at least one image information of the same goods, the goods are placed on a conveyor belt, and the conveyor belt moves a set distance within the acquisition interval time of the two continuous frames of image frames;
Obtaining the top point of the surface point cloud of the goods in each image frame, and obtaining the minimum bounding box of each goods;
Detecting a minimum bounding box with a corresponding relation in two frames of image frames based on the vertex and the minimum bounding box of the surface point cloud of the goods in the two frames of image frames;
And determining the target bounding box from the minimum bounding boxes with the corresponding relations.
In an alternative embodiment, the two image frames include a first image frame and a second image frame arranged according to the acquisition time sequence;
The step of detecting the minimum bounding box with the corresponding relation in the two frames of image frames based on the vertex and the minimum bounding box of the surface point cloud of the goods in the two frames of image frames comprises the following steps:
Screening out the minimum bounding box meeting the preset condition in the first image frame, and obtaining a key vertex in vertices corresponding to the screened minimum bounding box;
The moving direction of the conveyor belt is the x-axis direction of a camera coordinate system, the origin of the camera coordinate system is the center point of the camera visual field range of the image capturing device, the preset condition is that the x-coordinate value of the center point of the minimum bounding box is smaller than 0, and the minimum bounding box is not marked as a target bounding box;
And determining key vertexes of the minimum bounding boxes with corresponding relations based on the position information of the vertexes of the minimum bounding boxes and the position information of the key vertexes of the minimum bounding boxes screened in the first image frame aiming at the minimum bounding boxes in the second image frame so as to obtain the minimum bounding boxes with corresponding relations in the two image frames.
In an optional embodiment, the step of determining the key vertex of the minimum bounding box having the correspondence relationship based on the position information of each vertex of the minimum bounding box and the position information of the key vertex of each minimum bounding box screened in the first image frame includes:
determining target vertexes with minimum distances from the key vertexes based on the vertexes corresponding to the minimum bounding box and the position information of the key vertexes;
determining the distance from the minimum bounding box to each key vertex based on the target vertex of the minimum bounding box;
and detecting a minimum bounding box with a minimum distance from each key vertex, and determining that the key vertex with the minimum distance has a corresponding relation with the minimum bounding box.
In an optional embodiment, the step of obtaining the key vertex in the vertices corresponding to the screened minimum bounding box includes:
Aiming at each screened minimum bounding box, obtaining an x coordinate value of each vertex corresponding to the minimum bounding box;
Sequencing the vertexes according to the sequence from small to large of the x coordinate value;
And reversely traversing each vertex after sequencing to obtain a first vertex with the x coordinate value smaller than 0 as a key vertex.
In an optional embodiment, the step of determining the target bounding box from the smallest bounding boxes with the correspondence relationship includes:
Calculating the area of each minimum bounding box in the two minimum bounding boxes with the corresponding relation;
And determining the smallest bounding box with larger area of the two smallest bounding boxes with the corresponding relation as a target bounding box.
In an alternative embodiment, the step of obtaining vertices of a surface point cloud of the cargo in each of the image frames includes:
for each image frame, obtaining a surface point cloud of goods in the image frame, and projecting the surface point cloud to a two-dimensional coordinate system, wherein the two-dimensional coordinate system comprises a first coordinate axis and a second coordinate axis;
obtaining the minimum component and the maximum component of the projection point cloud projected on the first coordinate axis and the second coordinate axis respectively;
rotating the two-dimensional coordinate system, and obtaining the minimum component and the maximum component of the projection point cloud on the first coordinate axis and the second coordinate axis of the rotated two-dimensional coordinate system;
and obtaining the vertex of the surface point cloud based on the obtained multiple minimum components and maximum components.
In an alternative embodiment, the step of obtaining the minimum bounding box of each of the cargoes includes:
For each cargo, constructing a first rectangular edge based on two adjacent vertexes in the surface point cloud of the cargo, and constructing a second rectangular edge perpendicular to the first rectangular edge;
projecting each vertex in the surface point cloud to the directions of the first rectangular side and the second rectangular side respectively to obtain the minimum and maximum projection points in the directions of the first rectangular side and the second rectangular side respectively;
constructing a rectangle based on minimum and maximum projection points in the directions of the first rectangle side and the second rectangle side, and calculating the area of the rectangle;
And determining the rectangle with the smallest area in a plurality of rectangles obtained by the plurality of groups of adjacent two vertexes as the smallest bounding box.
In a second aspect, the present invention provides a bounding box extraction apparatus, the apparatus comprising:
the acquisition module is used for acquiring two continuous frames of image frames acquired by the camera equipment, wherein each image frame is provided with at least one piece of image information of the same goods, the goods are placed on the conveyor belt, and the conveyor belt moves for a set distance within the acquisition interval time of the two continuous frames of image frames;
the acquisition module is used for acquiring the vertexes of the surface point clouds of the cargoes in each image frame and acquiring the minimum bounding boxes of the cargoes;
the detection module is used for detecting the minimum bounding box with a corresponding relation in the two frames of image frames based on the vertexes of the surface point clouds of the cargoes in the two frames of image frames and the minimum bounding box;
And the determining module is used for determining the target bounding box from the minimum bounding boxes with the corresponding relations.
In a third aspect, the present invention provides an electronic device comprising one or more storage media and one or more processors in communication with the storage media, the one or more storage media storing machine-executable instructions that are executable by the processor to perform the method of any of the preceding embodiments when the electronic device is operating.
In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a method according to any of the preceding embodiments.
The beneficial effects of the embodiment of the invention include, for example:
The invention provides a bounding box extraction method, a bounding box extraction device, electronic equipment and a readable storage medium. And then obtaining the top points of the surface point clouds of the cargoes in each image frame, and obtaining the minimum bounding box of each cargoes. And detecting the minimum bounding box with the corresponding relation in the two frames of image frames based on the vertexes and the minimum bounding box of the surface point cloud in the two frames of image frames, and finally determining the target bounding box from the minimum bounding box with the corresponding relation. According to the scheme, the minimum bounding box with the corresponding relation in the two continuous image frames is detected, so that the target bounding box is determined, and the determined target bounding box can be ensured to be correct.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of minimum bounding box authenticity;
FIG. 2 (a) is one of the schematic diagrams of the minimum bounding box in the field of view of the camera;
FIG. 2 (b) is a second schematic diagram of a minimum bounding box in the field of view of the camera;
FIG. 3 is a flowchart of a bounding box extraction method provided by the present invention;
FIG. 4 is a schematic diagram of a camera coordinate system constructed in accordance with the present invention;
FIG. 5 is a flowchart of the substeps involved in S2 of FIG. 3;
FIG. 6 is another flow chart of sub-steps involved in S2 in FIG. 3;
FIG. 7 is a flowchart of the substeps involved in S3 of FIG. 3;
FIG. 8 is a flow chart of sub-steps involved in S31 of FIG. 7;
FIG. 9 is a schematic diagram of vertex realism;
FIG. 10 is a flow chart of sub-steps involved in S32 in FIG. 7;
FIG. 11 is a flowchart of the substeps involved in S4 of FIG. 3;
FIG. 12 is a functional block diagram of a bounding box extraction apparatus provided by the present invention;
Fig. 13 is a block diagram of an electronic device according to the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.
In the description of the present invention, it should be noted that, if the terms "first," "second," and the like are used merely to distinguish the descriptions, they are not to be construed as indicating or implying relative importance.
It should be noted that the features of the embodiments of the present invention may be combined with each other without conflict.
Referring to fig. 3, a flowchart of a bounding box extraction method provided by an embodiment of the present invention may be implemented by a bounding box extraction apparatus, which may be implemented by software and/or hardware, and may be configured in an electronic device, where the electronic device may be a computer device, a programmable logic controller, or the like. The detailed steps of the bounding box extraction method are described below.
S1, acquiring two continuous image frames acquired by an image pickup device, wherein each image frame has at least one image information of the same cargo.
The goods are placed on the conveyor belt, and the conveyor belt moves for a set distance within the acquisition interval time of two continuous image frames.
S2, obtaining the top points of the surface point clouds of the cargoes in each image frame, and obtaining the minimum bounding box of each cargoes.
And S3, detecting the minimum bounding box with a corresponding relation in the two frames of image frames based on the vertex of the surface point cloud of the goods in the two frames of image frames and the minimum bounding box.
S4, determining the target bounding box from the minimum bounding box with the corresponding relation.
In this embodiment, the electronic apparatus may be connected to the image capturing apparatus, and may be connected by wireless connection, wired connection, or the like. Wherein, the image pickup apparatus may be a depth camera, and the image pickup apparatus may be disposed above the conveyor belt with a shooting direction facing downward. The conveyor belt is within a camera field of view of the image pickup apparatus. When goods are placed on the conveyor belt, the goods and the conveyor belt are in the visual field of a camera for shooting.
In a practical application scenario, after the conveyor belt is started, the goods placed thereon are conveyed to pass under the image pickup apparatus in sequence. After each movement of the conveyor belt by a set distance, the image capturing apparatus performs image capturing once, that is, the conveyor belt moves by the set distance within the interval time of two consecutive image frames acquired by the image capturing apparatus.
In order to be able to perform determination of a true bounding box for each item, in the present embodiment, image information of at least one identical item is present in two consecutive image frames acquired by the image capturing apparatus. That is, each cargo appears at least 2 times in the camera field of view.
In the present embodiment, a camera coordinate system is constructed with the center of the camera field of view of the image pickup apparatus as the origin, and as shown in fig. 4, the direction in which the conveyor belt moves is the positive direction of the x-axis, and the y-axis is perpendicular to the direction in which the conveyor belt moves.
Generally, the length (dimension along the x-axis) of the cargo being targeted is less than or equal to L. Therefore, the length of the camera field of view of the image pickup apparatus can be set to 2L. After each L distance of the conveyor belt, the image pickup device performs image collection once. In this way, it is ensured that two consecutive image frames have image information of at least one and the same cargo, i.e. that each cargo appears at least 2 times in the field of view of the camera.
On the basis, for each of two consecutive image frames, a surface point cloud of the cargo in the image frame is obtained, and vertices of the surface point cloud are obtained. Wherein, the vertex of the surface point cloud of the goods can be determined based on the position information of each point, and can also be determined in other ways.
Furthermore, a minimum bounding box of each cargo is obtained, wherein the minimum bounding box refers to a minimum rectangle or cube that can completely enclose the cargo. In this embodiment, the analysis is mainly performed on the surface of the cargo, and thus, the minimum bounding box refers to the smallest rectangle that encloses the cargo.
From the above, the two consecutive image frames have at least one image information of the same cargo, so the minimum bounding box with a corresponding relationship in the two image frames can be determined based on the vertex and the minimum bounding box of the surface point cloud of the cargo in the two image frames, and the minimum bounding box with a corresponding relationship is the minimum bounding box corresponding to the same cargo.
In order to provide information for precisely grasping goods to an automation device, it is necessary to determine a target bounding box from among the minimum bounding boxes having a correspondence relationship, the target bounding box may be understood as a true minimum bounding box, and the other minimum bounding box may be understood as a false minimum bounding box. It can be seen that, in this embodiment, the minimum bounding box that is true and the minimum bounding box that is false are substantially a relative concept, that is, the minimum bounding box that is more beneficial to the automation device to accurately grasp the goods is true, and the other minimum bounding box that has a corresponding relationship is false.
In this embodiment, after the target bounding box is determined, information of the target bounding box may be provided to the automation device, so that the automation device may grasp the corresponding goods based on the target bounding box, and further perform stacking.
In the bounding box extraction scheme provided by the embodiment, two continuous image frames are acquired, and the conveyor belt moves a set distance within the acquisition interval time of the two image frames, wherein the two image frames have at least one image information of the same cargo. On the basis, the minimum bounding box with the corresponding relation in the two image frames is determined, so that the target bounding box is determined from the minimum bounding box with the corresponding relation. The scheme can ensure that the determined target bounding box is correct.
In this embodiment, after each frame of image frame acquired by the image capturing apparatus is obtained, the image frame may be processed to obtain the vertex of the surface point cloud of the cargo in the image frame. Referring to fig. 5, this step in this embodiment can be implemented by:
S21, for each image frame, obtaining a surface point cloud of the goods in the image frame, and projecting the surface point cloud to a two-dimensional coordinate system, wherein the two-dimensional coordinate system comprises a first coordinate axis and a second coordinate axis.
S22, obtaining the minimum component and the maximum component of the projection point cloud projected on the first coordinate axis and the second coordinate axis respectively.
S23, rotating the two-dimensional coordinate system, and obtaining the minimum component and the maximum component of the projection point cloud on the first coordinate axis and the second coordinate axis of the rotated two-dimensional coordinate system.
And S24, obtaining the vertex of the surface point cloud based on the obtained multiple minimum components and maximum components.
In the present embodiment, the obtained image frame is a depth image, which is a kind of gray-scale image, and the value of each pixel represents the distance of the pixel to the image pickup apparatus. The depth image may be converted into point cloud data using the depth value of each pixel in the depth image. In this way, a surface point cloud of the cargo in the image frame can be obtained. The surface point cloud refers to a point cloud of the cargo near the top plane of the image capturing apparatus.
Since in a real scenario the surface point cloud of the cargo may not be entirely on one plane, but rather be located near one plane. For convenience of processing, the surface point cloud is projected onto a constructed two-dimensional coordinate system to obtain a projected point cloud.
The two-dimensional coordinate system comprises a first coordinate axis and a second coordinate axis, wherein the first coordinate axis can be an x axis, and the second coordinate axis can be a y axis. In a two-dimensional coordinate system, each projection point has a coordinate value (x, y), where x represents a component of the projection point on a first coordinate axis and y represents a component of the projection point on a second coordinate axis. And respectively projecting each projection point in the projection point cloud to the first coordinate axis and the second coordinate axis. Thus, there are projected components on the first coordinate axis and projected components on the second coordinate axis. The minimum and maximum components having the minimum and maximum values on the first coordinate axis are obtained, and the minimum and maximum components having the minimum and maximum values on the second coordinate axis are obtained.
For the first coordinate axis, the minimum component and the maximum component on the first coordinate axis may be obtained, i.e. the component of the proxel on the first coordinate axis may be obtained but the component on the second coordinate axis is not determined. Therefore, in this embodiment, the second coordinate axis may be moved to the position of the minimum component and the maximum component of the first coordinate axis, where the movement is in a translational manner. And detecting the maximum value and the minimum value on the second coordinate axis at the moment, so that coordinate information of projection points corresponding to the minimum component and the maximum component in the projection point cloud in a two-dimensional coordinate system can be obtained.
In addition, for the second coordinate axis, the coordinate information of the projection point corresponding to the minimum component and the maximum component on the second coordinate axis in the projection point cloud in the two-dimensional coordinate system can be determined in the same manner.
By the above method, coordinate information of corresponding projection points of the two minimum components and the two maximum components in the projection point cloud can be obtained.
To determine a greater number of vertices, the two-dimensional coordinate system may also be rotated, for example, 45 degrees in either a clockwise or counterclockwise direction, in this embodiment. And obtaining the minimum component and the maximum component of the projection point cloud on the first coordinate axis and the second coordinate axis of the rotated two-dimensional coordinate system by adopting the same mode. And coordinate information of projection points corresponding to the minimum component and the maximum component in the projection point cloud is obtained.
In this embodiment, the number of rotations for the two-dimensional coordinate system may be one.
On the basis, a plurality of corresponding projection points in the projection point cloud can be obtained based on the obtained minimum components and maximum components, and each projection point can be used as the vertex of the surface point cloud of the goods. The vertices of the surface point cloud of the cargo may demarcate a boundary range of the surface of the cargo.
In this embodiment, after determining the vertex of the surface point cloud of the cargo, the minimum bounding box of each cargo may be obtained, and referring to fig. 6, the following steps may be implemented:
S25, for each cargo, constructing a first rectangular side based on two adjacent vertexes in the surface point cloud of the cargo, and constructing a second rectangular side perpendicular to the first rectangular side.
And S26, projecting each vertex in the surface point cloud to the directions of the first rectangular side and the second rectangular side respectively, and obtaining the minimum and maximum projection points in the directions of the first rectangular side and the second rectangular side respectively.
S27, constructing a rectangle based on minimum and maximum projection points in the directions of the first rectangle side and the second rectangle side, and calculating the area of the rectangle.
And S28, determining a rectangle with the smallest area in a plurality of rectangles obtained by a plurality of groups of adjacent two vertexes, and taking the rectangle as a smallest bounding box.
In this embodiment, for each vertex in the surface point cloud of the cargo, the vertices may be ordered in a certain order, for example, in a counterclockwise direction. And, each vertex is labeled 0,1,2 in order. For adjacent two of the vertices, e.g., 01, 12, etc., are not limited. A first rectangular edge may be constructed based on two adjacent vertices and a second rectangular edge connected to the first rectangular edge may be constructed in a direction perpendicular to the first rectangular edge.
Each vertex in the surface vertices can be projected towards the directions of the first rectangular edge and the second rectangular edge respectively, then the smallest projection point and the largest projection point on the first rectangular edge are obtained, and the smallest projection point and the largest projection point on the second rectangular edge are obtained. In this way, a rectangle can be constructed based on four proxels.
Rectangular construction is performed in the manner described above for every arbitrary two adjacent vertices in the surface point cloud, so that a plurality of rectangles can be obtained. The area of the rectangle can be calculated based on the length and width of each rectangle. Wherein the rectangle with the smallest area can be determined as the smallest bounding box of the cargo.
In this embodiment, two image frames may be divided into a first image frame and a second image frame according to the acquisition time sequence. On the basis of the above, the minimum bounding box with the corresponding relation in the two image frames is detected, referring to fig. 7, the steps can be implemented by the following ways:
And S31, screening out the minimum bounding box meeting the preset condition in the first image frame, and obtaining the key vertex in the vertexes corresponding to the screened minimum bounding box.
The moving direction of the conveyor belt is the x-axis direction of a camera coordinate system, the origin of the camera coordinate system is the center point of the camera visual field range of the image capturing device, the preset condition is that the x-coordinate value of the center point of the minimum bounding box is smaller than 0, and the minimum bounding box is not marked as a target bounding box.
S32, determining key vertexes of the minimum bounding boxes with corresponding relations according to the position information of the vertexes of the minimum bounding boxes and the position information of the key vertexes of the minimum bounding boxes screened in the first image frame, so as to obtain the minimum bounding boxes with corresponding relations in the two image frames.
In this embodiment, a minimum bounding box with an x coordinate value of the center point smaller than 0, i.e., in the left half of the camera field of view, is selected from the first image frame. In case the conveyor belt moves by a distance L each time, the length of the camera field of view is 2L, the maximum length of the minimum bounding box does not exceed L, this means that the minimum bounding box will also appear in the second image frame. In this way, a subsequent comparison with the corresponding smallest bounding box in the second image frame is made possible.
And for a minimum bounding box in the first image frame, the x coordinate value of the center point of the minimum bounding box is greater than or equal to 0, which indicates that the minimum bounding box has already occurred in the previous frame image of the first image frame, where the previous frame image frame can be considered as the first frame. In the latter detection, the second image frame is taken as the first image frame, and the image frame of the subsequent frame of the second image frame is taken as the second image frame. At this time, the minimum bounding box of the second image frame in the right half of the camera field of view has found the correspondence in the first image frame, and at least one of the minimum bounding boxes found the correspondence is true. Therefore, the minimum bounding box is not considered, and the minimum bounding box which is false is the minimum bounding box which needs to further determine the corresponding relation and make the true and false judgment. Therefore, in this embodiment, the minimum bounding box may also be defined and not marked as the target bounding box when screening in the first image frame.
And obtaining key vertexes corresponding to the minimum bounding boxes for the minimum bounding boxes which are screened out from the first image frames and meet the preset conditions. The key vertex is one of the vertices included in the surface point cloud.
Referring to fig. 8, in this embodiment, the step of obtaining the key vertex in the vertices corresponding to the minimum bounding box may be implemented as follows:
S311, obtaining the x coordinate values of the vertexes corresponding to the minimum bounding boxes aiming at the screened minimum bounding boxes.
S312, sorting the vertexes in the order of the x coordinate values from small to large.
S313, reversely traversing each vertex after sequencing to obtain a first vertex with x coordinate value smaller than 0 as a key vertex.
In this embodiment, the selected minimum bounding box may be marked as box_i, and the key vertex is substantially the maximum vertex whose x coordinate value is less than 0 in the vertices corresponding to the minimum bounding box. The purpose of determining the key vertices is to find the true vertices in the minimum bounding box, which vertices also occur in the second image frame. In order to satisfy that the vertex also appears in the second image frame, the x-coordinate value of the vertex needs to be less than 0. Wherein the true vertex means that the vertex is generated not as a result of the camera view intercepting the cargo. Therefore, the key vertex is defined as the largest (i.e. rightmost) vertex of all vertices with x coordinate values less than 0, so that the key vertex can be ensured to be a true vertex, i.e. the key vertex is not caused by capturing goods in the visual field of the camera. A schematic of a real vertex and a non-real vertex is shown in fig. 9.
In this embodiment, each vertex has a unique identifier of the index, and after determining the key vertex, the index of the key vertex may be recorded in the array box_i.counter_index_after determining the key vertex. After the minimum bounding box meeting the preset condition is screened out from the first image frame, adding an identification box_i of the minimum bounding box into an array box_need_cmp, wherein the box_need_cmp represents the minimum bounding box needing to determine the corresponding relation in the first image frame.
After determining the key vertex of the minimum bounding box with the corresponding relation in the first image frame, the minimum bounding box with the corresponding relation in the two frames can be detected based on the key vertex and the position information of the vertex of each minimum bounding box in the second image. Referring to fig. 10, this step may be implemented by:
s321, determining target vertexes with minimum distances from the key vertexes based on the vertexes corresponding to the minimum bounding box and the position information of the key vertexes aiming at the key vertexes.
S322, determining the distance from the minimum bounding box to each key vertex based on the target vertex of the minimum bounding box.
S323, for each key vertex, detecting a minimum bounding box with a minimum distance from the key vertex, and determining that the key vertex with the minimum distance has a corresponding relation with the minimum bounding box.
In this embodiment, for each minimum bounding box in the second image frame, the sorting may be performed in order from small to large based on the x coordinate value of the center point of the minimum bounding box. A processed array is initialized, the size of which matches the box_need_cmp size described above, indicating whether the corresponding minimum bounding box has been processed.
Let i=0, take the i-th minimum bounding box box_i in the second image frame, let j=0, take the j-th minimum bounding box box_j in box_need_cmp. The position information of the key vertex of the minimum bounding box box_j is obtained, and the sum value is obtained by adding the set distance moved by the conveyor belt to the position information of the key vertex, for example, L. And obtaining the position information of each vertex in the minimum bounding box box_i, calculating the difference value by utilizing the sum value and the position information of each vertex, and recording the calculated difference value into an array counter_dis_j.
Taking the vertex corresponding to the minimum difference value recorded in the counter_dis_j as a target vertex, taking the minimum difference value recorded in the counter_dis_j, and recording the minimum difference value into an array counter_dis.
Then j is increased until j exceeds the size of box_need_cmp, at which point the smallest difference recorded in the corer_dis represents the smallest distance from the ith smallest bounding box to the critical vertex of each smallest bounding box in box_need_cmp.
And taking the minimum distance in the counter_dis, judging whether the minimum distance is smaller than a preset value, and if so, judging that a corresponding relation exists between a key vertex corresponding to the minimum distance and a minimum bounding box, namely, that the minimum bounding box corresponding to the key vertex in the first image frame and the minimum bounding box in the second image frame have a corresponding relation.
At this time, the position min_pos_cd of the minimum bounding box in the first image frame corresponding to the minimum distance in the camera_dis may be obtained, and the minimum bounding box in the corresponding first image frame in the box_need_cmp [ min_pos_cd ] may be corresponding to the box_i, and the corresponding relationship may be recorded in the array box_i.id_pre. Further, setting processed [ min_pos_cd ] to 1 indicates that the minimum bounding box in the first image frame has found a correspondence.
And then increasing i until i exceeds the number of minimum bounding boxes in the second image frame.
After processing in the above manner, the correspondence between all the minimum bounding boxes in the first image frame and the second image frame can be determined. Since the minimum bounding box appearing in the first image frame also appears in the second image frame and is found according to the key vertices, the found correspondence is correct. And traversing the first image frame according to the sequence from small to large of the x coordinate value of the center point of the minimum bounding box in the second image frame, and arranging the corresponding minimum bounding boxes in the first image frame according to the sequence from small to large of the x coordinate value. And the minimum bounding box with a corresponding relation is marked every time a group of the minimum bounding boxes are found, so that the problem of repeated searching is avoided.
In this embodiment, after determining the minimum bounding box with the corresponding relationship in the first image frame and the second image frame, determining the target bounding box from the minimum bounding box with the corresponding relationship, referring to fig. 11, in this embodiment, the steps may be implemented by:
S41, calculating the area of each minimum bounding box in the two minimum bounding boxes with the corresponding relation.
S42, determining the smallest bounding box with larger area in the two smallest bounding boxes with the corresponding relation as a target bounding box.
Taking the above example as an illustration, after determining the minimum bounding box with the correspondence between the first image frame and the second image frame, the minimum bounding box in the second image frame may be traversed, and the ith second bounding box is set as box_i, and whether box_i.id_pre_is greater than or equal to 0 is determined, if greater than or equal to 0, it indicates that box_i has the corresponding minimum bounding box in the first image frame, and is denoted as box_j. And calculating the area of the box_i according to the length and the width of the box_i. In addition, the area of box_j is calculated from the length and width of box_j. The smallest bounding box corresponding to the larger area in the two is the target bounding box, namely the real bounding box.
Based on the same inventive concept, please refer to fig. 12, which is a schematic diagram illustrating functional modules of a bounding box extraction device according to an embodiment of the present invention, where the bounding box extraction device may be divided into functional modules according to the above-described method embodiment. For example, each functional module may be divided corresponding to each function, or two or more functions may be integrated in one processing module. The integrated modules may be implemented in hardware or in software functional modules. It should be noted that, in the embodiment of the present invention, the division of the modules is schematic, which is merely a logic function division, and other division manners may be implemented in actual implementation.
For example, in the case of dividing each function module by the corresponding each function, the bounding box extracting apparatus shown in fig. 12 is only one apparatus schematic. The bounding box extracting device may include an acquiring module, a detecting module and a determining module, and functions of each functional module of the bounding box extracting device are described in detail below.
The acquisition module is used for acquiring two continuous frames of image frames acquired by the camera equipment, wherein each image frame is provided with at least one piece of image information of the same goods, the goods are placed on the conveyor belt, and the conveyor belt moves for a set distance within the acquisition interval time of the two continuous frames of image frames;
It will be appreciated that the acquisition module may be used to perform S1 described above, and reference may be made to the details of S1 regarding the implementation of the acquisition module.
The acquisition module is used for acquiring the vertexes of the surface point clouds of the cargoes in each image frame and acquiring the minimum bounding boxes of the cargoes;
It will be appreciated that the acquisition module may be used to perform S2 described above, and reference may be made to the details of S2 regarding the implementation of the acquisition module.
The detection module is used for detecting the minimum bounding box with a corresponding relation in the two frames of image frames based on the vertexes of the surface point clouds of the cargoes in the two frames of image frames and the minimum bounding box;
It will be appreciated that the detection module may be used to perform S3 described above, and reference may be made to the details of S3 regarding the implementation of the detection module.
And the determining module is used for determining the target bounding box from the minimum bounding boxes with the corresponding relations.
It will be appreciated that the determination module may be used to perform S4 described above, and reference may be made to the details of S4 regarding the implementation of the determination module.
In one possible implementation manner, the two image frames include a first image frame and a second image frame arranged in sequence according to the acquisition time, and the detection module may be configured to:
Screening out the minimum bounding box meeting the preset condition in the first image frame, and obtaining a key vertex in vertices corresponding to the screened minimum bounding box;
The moving direction of the conveyor belt is the x-axis direction of a camera coordinate system, the origin of the camera coordinate system is the center point of the camera visual field range of the image capturing device, the preset condition is that the x-coordinate value of the center point of the minimum bounding box is smaller than 0, and the minimum bounding box is not marked as a target bounding box;
And determining key vertexes of the minimum bounding boxes with corresponding relations based on the position information of the vertexes of the minimum bounding boxes and the position information of the key vertexes of the minimum bounding boxes screened in the first image frame aiming at the minimum bounding boxes in the second image frame so as to obtain the minimum bounding boxes with corresponding relations in the two image frames.
In one possible implementation manner, the detection module may be used for:
determining target vertexes with minimum distances from the key vertexes based on the vertexes corresponding to the minimum bounding box and the position information of the key vertexes;
determining the distance from the minimum bounding box to each key vertex based on the target vertex of the minimum bounding box;
and detecting a minimum bounding box with a minimum distance from each key vertex, and determining that the key vertex with the minimum distance has a corresponding relation with the minimum bounding box.
In one possible implementation manner, the detection module may be used for:
Aiming at each screened minimum bounding box, obtaining an x coordinate value of each vertex corresponding to the minimum bounding box;
Sequencing the vertexes according to the sequence from small to large of the x coordinate value;
And reversely traversing each vertex after sequencing to obtain a first vertex with the x coordinate value smaller than 0 as a key vertex.
In one possible implementation manner, the determining module may be configured to:
Calculating the area of each minimum bounding box in the two minimum bounding boxes with the corresponding relation;
And determining the smallest bounding box with larger area of the two smallest bounding boxes with the corresponding relation as a target bounding box.
In one possible implementation manner, the obtaining module may be configured to:
for each image frame, obtaining a surface point cloud of goods in the image frame, and projecting the surface point cloud to a two-dimensional coordinate system, wherein the two-dimensional coordinate system comprises a first coordinate axis and a second coordinate axis;
obtaining the minimum component and the maximum component of the projection point cloud projected on the first coordinate axis and the second coordinate axis respectively;
rotating the two-dimensional coordinate system, and obtaining the minimum component and the maximum component of the projection point cloud on the first coordinate axis and the second coordinate axis of the rotated two-dimensional coordinate system;
and obtaining the vertex of the surface point cloud based on the obtained multiple minimum components and maximum components.
In one possible implementation manner, the obtaining module may be configured to:
For each cargo, constructing a first rectangular edge based on two adjacent vertexes in the surface point cloud of the cargo, and constructing a second rectangular edge perpendicular to the first rectangular edge;
projecting each vertex in the surface point cloud to the directions of the first rectangular side and the second rectangular side respectively to obtain the minimum and maximum projection points in the directions of the first rectangular side and the second rectangular side respectively;
constructing a rectangle based on minimum and maximum projection points in the directions of the first rectangle side and the second rectangle side, and calculating the area of the rectangle;
And determining the rectangle with the smallest area in a plurality of rectangles obtained by the plurality of groups of adjacent two vertexes as the smallest bounding box.
Referring to fig. 13, a block diagram of an electronic device according to an embodiment of the present invention may be a computer device, and the electronic device includes a memory, a processor, and a communication module. The memory, the processor and the communication module are electrically connected with each other directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines.
Wherein the memory is used for storing programs or data. The Memory may be, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), programmable Read Only Memory (Programmable Read-Only Memory, PROM), erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc.
The processor is used for reading/writing data or programs stored in the memory and executing the bounding box extraction method provided by any embodiment of the invention.
The communication module is used for establishing communication connection between the electronic equipment and other communication terminals through a network and is used for receiving and transmitting data through the network.
It should be understood that the configuration shown in fig. 13 is merely a schematic diagram of the configuration of the electronic device, and the electronic device may further include more or fewer components than those shown in fig. 13, or have a different configuration than that shown in fig. 13.
Further, an embodiment of the present invention further provides a computer readable storage medium storing machine executable instructions, where the machine executable instructions when executed implement the bounding box extraction method provided in the above embodiment.
In particular, the computer readable storage medium can be a general-purpose storage medium, such as a removable disk, a hard disk, or the like, and the computer program on the computer readable storage medium can perform the bounding box extraction method described above when executed. With respect to the processes in the computer readable storage medium and the executable instructions thereof involved when executed, reference is made to the relevant descriptions of the above method embodiments, which are not described in detail herein.
In summary, according to the bounding box extraction method, the bounding box extraction device, the electronic equipment and the readable storage medium provided by the embodiment of the invention, by acquiring two continuous frames of image frames acquired by the image pickup equipment, each image frame has at least one image information of the same goods, the goods are placed on the conveyor belt, and the conveyor belt moves a set distance within the acquisition interval time of the two continuous frames of image frames. And then obtaining the top points of the surface point clouds of the cargoes in each image frame, and obtaining the minimum bounding box of each cargoes. And detecting the minimum bounding box with the corresponding relation in the two frames of image frames based on the vertexes and the minimum bounding box of the surface point cloud in the two frames of image frames, and finally determining the target bounding box from the minimum bounding box with the corresponding relation. According to the scheme, the minimum bounding box with the corresponding relation in the two continuous image frames is detected, so that the target bounding box is determined, and the determined target bounding box can be ensured to be correct.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any changes or substitutions easily contemplated by those skilled in the art within the scope of the present invention should be included in the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A bounding box extraction method, the method comprising:
acquiring two continuous frames of image frames acquired by camera equipment, wherein each image frame has at least one image information of the same goods, the goods are placed on a conveyor belt, and the conveyor belt moves a set distance within the acquisition interval time of the two continuous frames of image frames;
Obtaining the top point of the surface point cloud of the goods in each image frame, and obtaining the minimum bounding box of each goods;
Detecting a minimum bounding box with a corresponding relation in two frames of image frames based on the vertex and the minimum bounding box of the surface point cloud of the goods in the two frames of image frames;
And determining the target bounding box from the minimum bounding boxes with the corresponding relations.
2. The bounding box extraction method of claim 1, wherein the two image frames include a first image frame and a second image frame arranged in chronological order of acquisition time;
The step of detecting the minimum bounding box with the corresponding relation in the two frames of image frames based on the vertex and the minimum bounding box of the surface point cloud of the goods in the two frames of image frames comprises the following steps:
Screening out the minimum bounding box meeting the preset condition in the first image frame, and obtaining a key vertex in vertices corresponding to the screened minimum bounding box;
The moving direction of the conveyor belt is the x-axis direction of a camera coordinate system, the origin of the camera coordinate system is the center point of the camera visual field range of the image capturing device, the preset condition is that the x-coordinate value of the center point of the minimum bounding box is smaller than 0, and the minimum bounding box is not marked as a target bounding box;
And determining key vertexes of the minimum bounding boxes with corresponding relations based on the position information of the vertexes of the minimum bounding boxes and the position information of the key vertexes of the minimum bounding boxes screened in the first image frame aiming at the minimum bounding boxes in the second image frame so as to obtain the minimum bounding boxes with corresponding relations in the two image frames.
3. The bounding box extraction method according to claim 2, wherein the step of determining the key vertices of the minimum bounding box having a correspondence relationship based on the positional information of the respective vertices of the minimum bounding box and the positional information of the key vertices of the respective minimum bounding boxes screened in the first image frame includes:
determining target vertexes with minimum distances from the key vertexes based on the vertexes corresponding to the minimum bounding box and the position information of the key vertexes;
determining the distance from the minimum bounding box to each key vertex based on the target vertex of the minimum bounding box;
and detecting a minimum bounding box with a minimum distance from each key vertex, and determining that the key vertex with the minimum distance has a corresponding relation with the minimum bounding box.
4. The bounding box extraction method according to claim 2, wherein the step of obtaining the key vertex among the vertices corresponding to the screened minimum bounding box comprises:
Aiming at each screened minimum bounding box, obtaining an x coordinate value of each vertex corresponding to the minimum bounding box;
Sequencing the vertexes according to the sequence from small to large of the x coordinate value;
And reversely traversing each vertex after sequencing to obtain a first vertex with the x coordinate value smaller than 0 as a key vertex.
5. The bounding box extraction method according to claim 1, wherein the step of determining the target bounding box from the smallest bounding box having the correspondence relationship comprises:
Calculating the area of each minimum bounding box in the two minimum bounding boxes with the corresponding relation;
And determining the smallest bounding box with larger area of the two smallest bounding boxes with the corresponding relation as a target bounding box.
6. The bounding box extraction method of claim 1, wherein the step of obtaining vertices of a surface point cloud of the cargo in each of the image frames comprises:
for each image frame, obtaining a surface point cloud of goods in the image frame, and projecting the surface point cloud to a two-dimensional coordinate system, wherein the two-dimensional coordinate system comprises a first coordinate axis and a second coordinate axis;
obtaining the minimum component and the maximum component of the projection point cloud projected on the first coordinate axis and the second coordinate axis respectively;
rotating the two-dimensional coordinate system, and obtaining the minimum component and the maximum component of the projection point cloud on the first coordinate axis and the second coordinate axis of the rotated two-dimensional coordinate system;
and obtaining the vertex of the surface point cloud based on the obtained multiple minimum components and maximum components.
7. The bounding box extraction method of claim 1, wherein the step of obtaining a minimum bounding box for each of the goods comprises:
For each cargo, constructing a first rectangular edge based on two adjacent vertexes in the surface point cloud of the cargo, and constructing a second rectangular edge perpendicular to the first rectangular edge;
projecting each vertex in the surface point cloud to the directions of the first rectangular side and the second rectangular side respectively to obtain the minimum and maximum projection points in the directions of the first rectangular side and the second rectangular side respectively;
constructing a rectangle based on minimum and maximum projection points in the directions of the first rectangle side and the second rectangle side, and calculating the area of the rectangle;
And determining the rectangle with the smallest area in a plurality of rectangles obtained by the plurality of groups of adjacent two vertexes as the smallest bounding box.
8. A bounding box extraction apparatus, the apparatus comprising:
the acquisition module is used for acquiring two continuous frames of image frames acquired by the camera equipment, wherein each image frame is provided with at least one piece of image information of the same goods, the goods are placed on the conveyor belt, and the conveyor belt moves for a set distance within the acquisition interval time of the two continuous frames of image frames;
the acquisition module is used for acquiring the vertexes of the surface point clouds of the cargoes in each image frame and acquiring the minimum bounding boxes of the cargoes;
the detection module is used for detecting the minimum bounding box with a corresponding relation in the two frames of image frames based on the vertexes of the surface point clouds of the cargoes in the two frames of image frames and the minimum bounding box;
And the determining module is used for determining the target bounding box from the minimum bounding boxes with the corresponding relations.
9. An electronic device comprising one or more storage media and one or more processors in communication with the storage media, the one or more storage media storing processor-executable machine-executable instructions that, when the electronic device is run, are executed by the processor to perform the method of any of claims 1-7.
10. A computer readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410642952.7A CN118247473B (en) | 2024-05-23 | 2024-05-23 | Bounding box extraction method, device, electronic equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410642952.7A CN118247473B (en) | 2024-05-23 | 2024-05-23 | Bounding box extraction method, device, electronic equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118247473A true CN118247473A (en) | 2024-06-25 |
| CN118247473B CN118247473B (en) | 2024-09-27 |
Family
ID=91557563
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410642952.7A Active CN118247473B (en) | 2024-05-23 | 2024-05-23 | Bounding box extraction method, device, electronic equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118247473B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106000904A (en) * | 2016-05-26 | 2016-10-12 | 北京新长征天高智机科技有限公司 | Automatic sorting system for household refuse |
| CN110276317A (en) * | 2019-06-26 | 2019-09-24 | Oppo广东移动通信有限公司 | A kind of dimension of object detection method, dimension of object detection device and mobile terminal |
| CN112150533A (en) * | 2019-06-28 | 2020-12-29 | 顺丰科技有限公司 | Object volume calculation method, device, equipment and storage medium |
| CN113450459A (en) * | 2020-03-25 | 2021-09-28 | 北京四维图新科技股份有限公司 | Method and device for constructing three-dimensional model of target object |
| CN117994561A (en) * | 2023-12-26 | 2024-05-07 | 清华大学 | Target detection method and device based on directed bounding box and electronic equipment |
-
2024
- 2024-05-23 CN CN202410642952.7A patent/CN118247473B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106000904A (en) * | 2016-05-26 | 2016-10-12 | 北京新长征天高智机科技有限公司 | Automatic sorting system for household refuse |
| CN110276317A (en) * | 2019-06-26 | 2019-09-24 | Oppo广东移动通信有限公司 | A kind of dimension of object detection method, dimension of object detection device and mobile terminal |
| CN112150533A (en) * | 2019-06-28 | 2020-12-29 | 顺丰科技有限公司 | Object volume calculation method, device, equipment and storage medium |
| CN113450459A (en) * | 2020-03-25 | 2021-09-28 | 北京四维图新科技股份有限公司 | Method and device for constructing three-dimensional model of target object |
| CN117994561A (en) * | 2023-12-26 | 2024-05-07 | 清华大学 | Target detection method and device based on directed bounding box and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118247473B (en) | 2024-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1678659B1 (en) | Method and image processing device for analyzing an object contour image, method and image processing device for detecting an object, industrial vision apparatus, smart camera, image display, security system, and computer program product | |
| US10445868B2 (en) | Method for detecting a defect on a surface of a tire | |
| US11232589B2 (en) | Object recognition device and object recognition method | |
| CN113962274B (en) | Abnormity identification method and device, electronic equipment and storage medium | |
| WO2021249406A1 (en) | Cargo box extraction and device, system, robot, and storage medium | |
| AU2011290557B2 (en) | Sensor data processing | |
| Sansoni et al. | Optoranger: A 3D pattern matching method for bin picking applications | |
| CN117190866B (en) | Polarity discrimination detection method, device and equipment for multiple stacked electronic components | |
| EP2458556A1 (en) | Marker generation device, marker generation detection system, marker generation detection device, marker, marker generation method, and program therefor | |
| Holz et al. | Fast edge-based detection and localization of transport boxes and pallets in rgb-d images for mobile robot bin picking | |
| CN110673607A (en) | Feature point extraction method and device in dynamic scene and terminal equipment | |
| CN112489240B (en) | Commodity display inspection method, inspection robot and storage medium | |
| CN118247473B (en) | Bounding box extraction method, device, electronic equipment and readable storage medium | |
| CN115170791A (en) | Target object identification method and device and vending equipment | |
| CN110211161B (en) | Method and device for multi-camera track coincidence analysis | |
| US11506483B2 (en) | Method, system and apparatus for support structure depth determination | |
| US20160205395A1 (en) | Method for detecting errors for at least one image processing system | |
| CN114882058B (en) | Corner detection method, corner detection device and calibration plate | |
| CN116342858B (en) | Object detection method, device, electronic equipment and storage medium | |
| CN112614181B (en) | Robot positioning method and device based on highlight target | |
| CN113255737B (en) | Method for sorting baggage in folded package on civil aviation sorting line, electronic equipment and storage medium | |
| CN118172526A (en) | Target positioning method, device, equipment and storage medium | |
| US20160205396A1 (en) | Method for error detection for at least one image processing system | |
| Garnaik et al. | MINSU: Precision Quantity Counter with DNN-based Volume Estimation | |
| CN112528722A (en) | Book checking method and system based on SSD-ResNet algorithm |
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 |