CN114170373A - Target object labeling method, processor, device and mixing station - Google Patents
Target object labeling method, processor, device and mixing station Download PDFInfo
- Publication number
- CN114170373A CN114170373A CN202111228911.6A CN202111228911A CN114170373A CN 114170373 A CN114170373 A CN 114170373A CN 202111228911 A CN202111228911 A CN 202111228911A CN 114170373 A CN114170373 A CN 114170373A
- Authority
- CN
- China
- Prior art keywords
- target object
- image
- determining
- semantic
- image acquisition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
Abstract
The embodiment of the application provides a target object labeling method, a processor, a device and a mixing station. The method comprises the following steps: obtaining an image to be marked of a target object through image acquisition equipment; establishing a corresponding world coordinate system according to the installation position of the image acquisition equipment; carrying out space projection on an image to be marked in a world coordinate system; determining a corresponding space coordinate of each pixel in a world coordinate system, wherein the pixel is contained in an image to be annotated; performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to each space coordinate; mapping the space coordinates to an image to be marked to obtain corresponding plane coordinates; determining a second semantic label of the plane coordinate according to the first semantic label; and determining the area of the target object in the image to be labeled according to the second semantic label. By the method, an automatic labeling process is realized, the period of data labeling is greatly shortened, the time and labor cost are effectively saved, and the labeling efficiency is effectively improved.
Description
Technical Field
The application relates to the technical field of computers, in particular to a target object labeling method, a processor, a device and a mixing station.
Background
In the working process of the engineering machinery, a plurality of tasks can be modeled as deep learning semantic segmentation tasks, and automatic operation is carried out by utilizing an algorithm to replace manual work, so that the productivity is improved. In this process, labeling of semantic segmentation is a necessary but very complicated and time-consuming task, and the labeling personnel needs to label the type of each pixel. This labeling at the pixel level is a very laborious and laborious process. However, in the semantic segmentation task of the construction machine, the labeling task is a work that must be completed. In the prior art, manual labeling is usually performed through labeling software, and the process can be summarized as follows: and selecting a marked object picture, loading the picture into marking software, manually marking, and adding a data set after marking for subsequent model training. However, this purely manual labeling approach is not only time consuming and labor intensive, but also has a low input-output ratio.
Disclosure of Invention
The embodiment of the application aims to provide a target object labeling method, a processor, a device and a mixing station, wherein the time and labor cost are saved, and the labeling efficiency can be improved.
In order to achieve the above object, a first aspect of the present application provides a target object labeling method, including:
obtaining an image to be marked of a target object through image acquisition equipment;
establishing a corresponding world coordinate system according to the installation position of the image acquisition equipment;
carrying out space projection on an image to be marked in a world coordinate system;
determining a corresponding space coordinate of each pixel in a world coordinate system, wherein the pixel is contained in an image to be annotated;
performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to each space coordinate;
mapping the space coordinates to an image to be marked to obtain corresponding plane coordinates;
determining a second semantic label of the plane coordinate according to the first semantic label;
and determining the area of the target object in the image to be labeled according to the second semantic label.
In an embodiment of the present application, performing semantic segmentation on the spatial coordinates, and determining a first semantic tag corresponding to each spatial coordinate includes: randomly selecting two space coordinates from the space coordinates, and determining the distance between the two space coordinates; the spatial coordinates are classified according to distance to determine a first semantic tag for each spatial coordinate.
In an embodiment of the present application, determining, according to the second semantic label, a region where the target object is located in the image to be annotated includes: and determining the area formed by the plane coordinates with the same second semantic label as the area of the target object in the image to be annotated.
In an embodiment of the present application, further comprising: before the image to be marked of the target object is obtained through the image acquisition equipment, camera calibration is carried out on the image acquisition equipment to determine camera parameters, and frame alignment operation is carried out.
In the embodiment of the application, the number of the image acquisition equipment is multiple, so that the target object can be shot in multiple angles.
In an embodiment of the present application, determining, according to the second semantic label, a region where the target object is located in the image to be annotated includes: determining an image acquisition area of each image acquisition device; determining a vacuum area of the image acquisition equipment according to the image acquisition area, wherein the vacuum area is an area which cannot be shot by all the image acquisition equipment; determining a third semantic label corresponding to the vacuum area; and determining the area of the target object in the image to be labeled according to the second semantic label and the third semantic label.
In the embodiment of the present application, when an image to be annotated contains a plurality of target objects, performing semantic segmentation on spatial coordinates, and determining a first semantic label corresponding to each spatial coordinate includes: and performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to the space coordinates of each target object, wherein the first semantic labels corresponding to the space coordinates of each target object are the same.
In the embodiment of the application, the target object labeling method is applied to the mixing station to identify the feeding hole of the mixing station.
A second aspect of the present application provides a processor configured to execute the target object labeling method described above.
A third aspect of the present application provides a target object labeling apparatus, including the processor described above.
The present application fourth aspect provides a mixing plant, the mixing plant comprising:
the image acquisition equipment is used for acquiring an image to be annotated of the target object;
a target object; and
the target object labeling device is described above.
According to the target object labeling method, the image to be labeled of the target object is acquired through the installed image acquisition equipment, the image is projected to the three-dimensional space through parallax estimation, the space coordinate corresponding to each pixel is determined, then the object points are automatically segmented in the space through an algorithm, and the space coordinate with the label is projected back to the original 2D image to obtain the pixel point label of the original 2D image.
Additional features and advantages of embodiments of the present application will be described in detail in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the embodiments of the disclosure, but are not intended to limit the embodiments of the disclosure. In the drawings:
FIG. 1 is a diagram schematically illustrating an application environment of a target object labeling method according to an embodiment of the present application;
FIG. 2 is a schematic flow chart illustrating a target object labeling method according to an embodiment of the present application;
FIG. 3 schematically illustrates a block diagram of a mixing station according to an embodiment of the present application;
fig. 4 schematically shows an internal structure diagram of a computer device according to an embodiment of the present application.
Detailed Description
The following detailed description of embodiments of the present application will be made with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the embodiments of the application, are given by way of illustration and explanation only, not limitation.
The target object labeling method provided by the application can be applied to the application environment shown in fig. 1. The image capturing device may include a plurality of devices, such as the image capturing device 102, the image capturing device 104, and the like. The image capture device may communicate with the processor 106 over a network. The image to be annotated of the target object can be acquired by the image acquisition devices 102 and 104, and the like, and the image to be annotated is transmitted to the processor 106, so that the processor 106 determines the area of the target object in the image to be annotated. The image capturing devices 102 and 104 may be, but not limited to, various devices with image capturing functions, such as a video camera, a still camera, a recorder, a notebook computer, a smart phone, a tablet computer, and a portable wearable device with a camera function, and the processor 106 may be implemented by an independent server or a server cluster formed by a plurality of servers.
Fig. 2 schematically shows a flow chart of a target object labeling method according to an embodiment of the present application. As shown in fig. 2, in an embodiment of the present application, a target object labeling method is provided, which includes the following steps:
And step 203, performing spatial projection on the image to be annotated in a world coordinate system.
And 204, determining the corresponding space coordinate of each pixel in the image to be annotated in the world coordinate system.
And step 206, mapping the space coordinates to the image to be annotated to obtain corresponding plane coordinates.
And step 207, determining a second semantic label of the plane coordinate according to the first semantic label.
And step 208, determining the area of the target object in the image to be labeled according to the second semantic label.
Firstly, image acquisition equipment can be installed, the number of the image acquisition equipment can be multiple, so that the image acquisition equipment can be used for carrying out image acquisition on a target object at multiple angles, and the image acquisition equipment can be used for carrying out image acquisition on the area where the target object is located to obtain a corresponding image to be annotated. Specifically, the image capturing device may be installed obliquely above the mixing station for identifying the feed opening of the mixing station. Further, a corresponding world coordinate system may be established according to the installation location of the image capturing device. For example, if there are 3 image capturing devices, a world coordinate system corresponding to the image capturing devices may be established according to the installation locations of the three image capturing devices, so that all the three image capturing devices may be included in the world coordinate system. Then, the image to be annotated can be subjected to space projection in a world coordinate system. Specifically, the disparity algorithm may be used to perform spatial three-dimensional projection on the image to be annotated, that is, three-dimensional spatial point coordinates of the image to be annotated are calculated, so as to determine spatial coordinates corresponding to each pixel included in the image to be annotated in a world coordinate system. Furthermore, semantic segmentation can be performed on each spatial coordinate, and a first semantic label corresponding to each spatial coordinate is determined.
In one embodiment, performing semantic segmentation on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate includes: randomly selecting two space coordinates from the space coordinates, and determining the distance between the two space coordinates; the spatial coordinates are classified according to distance to determine a first semantic tag for each spatial coordinate.
Each pixel in the image to be marked has a corresponding space coordinate, two space coordinates can be randomly selected from all the space coordinates, the distance between the two selected space coordinates is calculated, and the space coordinates can be classified according to the distance. Specifically, a distance threshold may be set, and when the distance between two spatial coordinates is smaller than the distance threshold, the two spatial coordinates may be considered to belong to the same type, and then the first semantic tags corresponding to the two spatial coordinates may be set as the same tags. For example, corresponding tag 1 is added for both tags.
After the first semantic label corresponding to each spatial coordinate is determined, the spatial coordinate can be mapped to an image to be annotated, and since the image to be annotated is a two-dimensional image, a plane coordinate corresponding to each spatial coordinate, that is, a plane coordinate of each 2D pixel point in the image to be annotated, can be obtained after the spatial coordinate is mapped to the image to be annotated. And then, determining a second semantic label of the plane coordinate according to the first semantic label, namely mapping the first semantic label of the space coordinate to a 2D plane according to the first semantic label of the space coordinate to obtain the second semantic label of the plane coordinate, so as to complete an automatic labeling process aiming at each pixel point. And then, the processor can automatically determine the area of the target object in the image to be labeled according to the second semantic tag. In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: and determining the area formed by the plane coordinates with the same second semantic label as the area of the target object in the image to be annotated. For example, assuming that the first semantic tag is 1 and represents that the target object is included, when the second semantic tag of the plane coordinates is determined according to the first semantic tag, the coordinate point where the second semantic tag is also 1 may be determined as the coordinate point including the target object, and then the area where the target object is located may be determined according to all the plane coordinates where the tags are 1. Assuming that there are 50 coordinate points with the second semantic label of 1, the region composed of these 50 coordinate points may be determined as the region where the target object is located in the image to be labeled. The annotated data can subsequently be saved to a database for subsequent use.
In one embodiment, the method further comprises: before the image to be marked of the target object is obtained through the image acquisition equipment, camera calibration is carried out on the image acquisition equipment to determine camera parameters, and frame alignment operation is carried out.
The image capturing device may be calibrated prior to use. Typically, the image capture device will be mounted in a relatively fixed position. For example, the image capturing device may be installed obliquely above the feed port of the construction machine. The work machine may be a mixing plant. At this time, the image capturing device at a fixed position may be calibrated, for example, a gnomon calibration method may be adopted. The process of calibration can be summarized in two parts:
the first step is to convert the world coordinate system into a camera coordinate system, and the first step is to convert three-dimensional points into three-dimensional points, wherein the three-dimensional points comprise parameters such as R RR, tt (camera external reference) and the like;
the second step is to convert the camera coordinate system into an image coordinate system, and the step is to convert a three-dimensional point into a two-dimensional point, and comprises parameters such as K KK (camera reference).
After video frames of the image acquisition device are aligned, camera internal parameters of the image acquisition device can be determined. The camera internal parameters of each image capturing device may have certain differences, which may be specific to the specific situation of each image capturing device. In one embodiment, the number of the image capturing devices may be multiple, so that the target objects at multiple angles can be captured, and the range of the blind area can be reduced if the images captured by the multiple image capturing devices are combined. For example, if the image capturing devices are few, the target object may not be captured completely, or the target object at some angles may not be captured clearly, and so on.
In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: determining an image acquisition area of each image acquisition device; determining a vacuum area of the image acquisition equipment according to the image acquisition area, wherein the vacuum area is an area which cannot be shot by all the image acquisition equipment; determining a third semantic label corresponding to the vacuum area; and determining the area of the target object in the image to be labeled according to the second semantic label and the third semantic label.
Under the condition that a plurality of image acquisition devices are provided, the image acquisition region of each image acquisition device can be acquired, and then the vacuum region of each image acquisition device is determined. The vacuum region is a region that cannot be captured by all image capturing apparatuses. For example, the large area in which the target object is located may be divided into ten small areas Q1-Q10. And all image acquisition devices cannot shoot the areas Q6 and Q7, so the areas Q6 and Q7 are vacuum areas. In this case, a third semantic tag corresponding to the vacuum region may be specified, for example, a tag 0 may be specified as a tag corresponding to the vacuum region, and if the semantic tag is specified as 0, it indicates that the region belongs to the vacuum region. And determining the region of the target object in the image to be marked by combining the second semantic label of the plane coordinate and the third semantic label of the vacuum region. The advantage of this process is that it can be further determined whether there is a portion of the second semantic label that coincides with the third semantic label, for example, the second semantic label of the coordinate point X1 is 1, but the coordinate point X1 is determined as a vacuum region and is labeled with 0, at this time, the semantic label of the coordinate point X1 can be corrected, so that the region of the target object in the image to be labeled can be determined more accurately. Further, a background label may also be introduced. If the target object is overlapped with other objects, the target area and the background object where the target area is located need to be distinguished, then the edge point of the target object needs to be accurately determined, so that the outline of the target object can be determined, and the area where the target object is located in the image to be marked can be more accurately determined. For example, assume that user A is standing in front of a building. The user a is a target object and needs to determine the area of the user a in the picture. Then, the building in the picture is the background object, and the background object needs to be marked out, so that the outline of the user a can be accurately determined. Preferably, to improve the intelligent operation of the process, a neural network can be introduced to determine the target object and the background object in the same picture. To improve the accuracy of this procedure, manual intervention may also be used for the marking. It should be added that, in the embodiments of the present application, how to label the target object is considered, and it is not necessary to consider what the target object is specifically.
According to the target object labeling method, the image to be labeled of the target object is acquired through the installed image acquisition equipment, the image is projected to the three-dimensional space through parallax estimation, the space coordinate corresponding to each pixel is determined, then the object points are automatically segmented in the space through an algorithm, and the space coordinate with the label is projected back to the original 2D image to obtain the pixel point label of the original 2D image.
An embodiment of the present application provides a storage medium, on which a program is stored, and the program, when executed by a processor, implements the above target object labeling method.
The embodiment of the application provides a processor, wherein the processor is used for running a program, and the target object labeling method is executed when the program runs.
In one embodiment, a target object labeling apparatus is provided, which includes the processor described above.
The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. One or more than one kernel can be set, and the target object labeling method is realized by adjusting kernel parameters.
In one embodiment, as shown in fig. 3, there is provided a mixing station 300 comprising:
the image acquisition equipment 301 is used for acquiring an image to be annotated of a target object;
a target object 302; and
target object labeling means 303.
Among them, the number of the image pickup devices 301 may be plural. The target object may be referred to as a feed inlet of the mixing station.
The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 4. The computer device includes a processor a01, a network interface a02, a memory (not shown), and a database (not shown) connected by a system bus. Wherein processor a01 of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises an internal memory a03 and a non-volatile storage medium a 04. The non-volatile storage medium a04 stores an operating system B01, a computer program B02, and a database (not shown in the figure). The internal memory a03 provides an environment for the operation of the operating system B01 and the computer program B02 in the nonvolatile storage medium a 04. The database of the computer equipment is used for storing data such as images to be marked. The network interface a02 of the computer device is used for communication with an external terminal through a network connection. The computer program B02 is adapted to carry out a method of target object annotation when executed by the processor a 01.
Those skilled in the art will appreciate that the architecture shown in fig. 4 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
The embodiment of the application provides equipment, the equipment comprises a processor, a memory and a program which is stored on the memory and can run on the processor, and the following steps are realized when the processor executes the program: obtaining an image to be marked of a target object through image acquisition equipment; establishing a corresponding world coordinate system according to the installation position of the image acquisition equipment; carrying out space projection on an image to be marked in a world coordinate system; determining a corresponding space coordinate of each pixel in a world coordinate system, wherein the pixel is contained in an image to be annotated; performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to each space coordinate; mapping the space coordinates to an image to be marked to obtain corresponding plane coordinates; determining a second semantic label of the plane coordinate according to the first semantic label; and determining the area of the target object in the image to be labeled according to the second semantic label.
In one embodiment, performing semantic segmentation on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate includes: randomly selecting two space coordinates from the space coordinates, and determining the distance between the two space coordinates; the spatial coordinates are classified according to distance to determine a first semantic tag for each spatial coordinate.
In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: and determining the area formed by the plane coordinates with the same second semantic label as the area of the target object in the image to be annotated.
In one embodiment, the steps further comprise: before the image to be marked of the target object is obtained through the image acquisition equipment, camera calibration is carried out on the image acquisition equipment to determine camera parameters, and frame alignment operation is carried out.
In one embodiment, the number of the image acquisition devices is multiple, so that the target object can be shot in multiple angles.
In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: determining an image acquisition area of each image acquisition device; determining a vacuum area of the image acquisition equipment according to the image acquisition area, wherein the vacuum area is an area which cannot be shot by all the image acquisition equipment; determining a third semantic label corresponding to the vacuum area; and determining the area of the target object in the image to be labeled according to the second semantic label and the third semantic label.
In one embodiment, when the image to be annotated contains a plurality of target objects, performing semantic segmentation on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate includes: and performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to the space coordinates of each target object, wherein the first semantic labels corresponding to the space coordinates of each target object are the same.
In one embodiment, the target object labeling method is applied to a mixing station to identify a feed inlet of the mixing station.
The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device: obtaining an image to be marked of a target object through image acquisition equipment; establishing a corresponding world coordinate system according to the installation position of the image acquisition equipment; carrying out space projection on an image to be marked in a world coordinate system; determining a corresponding space coordinate of each pixel in a world coordinate system, wherein the pixel is contained in an image to be annotated; performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to each space coordinate; mapping the space coordinates to an image to be marked to obtain corresponding plane coordinates; determining a second semantic label of the plane coordinate according to the first semantic label; and determining the area of the target object in the image to be labeled according to the second semantic label.
In one embodiment, performing semantic segmentation on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate includes: randomly selecting two space coordinates from the space coordinates, and determining the distance between the two space coordinates; the spatial coordinates are classified according to distance to determine a first semantic tag for each spatial coordinate.
In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: and determining the area formed by the plane coordinates with the same second semantic label as the area of the target object in the image to be annotated.
In one embodiment, the steps further comprise: before the image to be marked of the target object is obtained through the image acquisition equipment, camera calibration is carried out on the image acquisition equipment to determine camera parameters, and frame alignment operation is carried out.
In one embodiment, the number of the image acquisition devices is multiple, so that the target object can be shot in multiple angles.
In one embodiment, determining the region of the target object in the image to be annotated according to the second semantic label comprises: determining an image acquisition area of each image acquisition device; determining a vacuum area of the image acquisition equipment according to the image acquisition area, wherein the vacuum area is an area which cannot be shot by all the image acquisition equipment; determining a third semantic label corresponding to the vacuum area; and determining the area of the target object in the image to be labeled according to the second semantic label and the third semantic label.
In one embodiment, when the image to be annotated contains a plurality of target objects, performing semantic segmentation on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate includes: and performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to the space coordinates of each target object, wherein the first semantic labels corresponding to the space coordinates of each target object are the same.
In one embodiment, the target object labeling method is applied to a mixing station to identify a feed inlet of the mixing station.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.
The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (12)
1. A target object labeling method is characterized by comprising the following steps:
obtaining an image to be marked of a target object through image acquisition equipment;
establishing a corresponding world coordinate system according to the installation position of the image acquisition equipment;
performing spatial projection on the image to be marked in the world coordinate system;
determining a corresponding space coordinate of each pixel in the to-be-annotated image in the world coordinate system;
performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to each space coordinate;
mapping the space coordinate to the image to be marked to obtain a corresponding plane coordinate;
determining a second semantic label of the plane coordinate according to the first semantic label;
and determining the area of the target object in the image to be annotated according to the second semantic label.
2. The method for labeling a target object according to claim 1, wherein the semantic segmentation is performed on the spatial coordinates, and determining the first semantic tag corresponding to each spatial coordinate comprises:
randomly selecting two space coordinates from the space coordinates, and determining the distance between the two space coordinates;
and classifying the space coordinates according to the distance to determine a first semantic label of each space coordinate.
3. The method for labeling a target object according to claim 1, wherein the determining the region of the target object in the image to be labeled according to the second semantic tag comprises:
and determining a region formed by the plane coordinates with the same second semantic label as the region of the target object in the image to be annotated.
4. The target object labeling method of claim 1, further comprising:
before the image to be marked of the target object is obtained through the image acquisition equipment, calibrating a camera of the image acquisition equipment to determine camera parameters, and performing frame alignment operation.
5. The method for labeling a target object according to claim 1, wherein the number of the image capturing devices is plural for capturing the target object from multiple angles.
6. The method for labeling a target object according to claim 5, wherein the determining the region of the target object in the image to be labeled according to the second semantic tag comprises:
determining an image acquisition area of each image acquisition device;
determining a vacuum area of the image acquisition equipment according to the image acquisition area, wherein the vacuum area is an area which cannot be shot by all the image acquisition equipment;
determining a third semantic label corresponding to the vacuum region;
and determining the area of the target object in the image to be annotated according to the second semantic label and the third semantic label.
7. The method for labeling a target object according to claim 1, wherein, when the image to be labeled contains a plurality of target objects, the semantically segmenting the spatial coordinates, and determining the first semantic label corresponding to each spatial coordinate comprises:
and performing semantic segmentation on the space coordinates, and determining a first semantic label corresponding to the space coordinates of each target object, wherein the first semantic labels corresponding to the space coordinates of each target object are the same.
8. The target object labeling method of claim 1, applied to a mixing station to identify a feed inlet of the mixing station.
9. A processor configured to perform the target object labeling method of any one of claims 1 to 8.
10. A target object annotation device comprising a processor according to claim 9.
11. A mixing station, characterized in that it comprises:
the image acquisition equipment is used for acquiring an image to be annotated of the target object;
a target object; and
the target object annotation device of claim 10.
12. The mixing station of claim 11, wherein the target object is a feed gap.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111228911.6A CN114170373A (en) | 2021-10-21 | 2021-10-21 | Target object labeling method, processor, device and mixing station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111228911.6A CN114170373A (en) | 2021-10-21 | 2021-10-21 | Target object labeling method, processor, device and mixing station |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114170373A true CN114170373A (en) | 2022-03-11 |
Family
ID=80477148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111228911.6A Pending CN114170373A (en) | 2021-10-21 | 2021-10-21 | Target object labeling method, processor, device and mixing station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114170373A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116704008A (en) * | 2023-08-01 | 2023-09-05 | 城云科技（中国）有限公司 | Method and device for judging object based on picture area calculation and application of method and device |
-
2021
- 2021-10-21 CN CN202111228911.6A patent/CN114170373A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116704008A (en) * | 2023-08-01 | 2023-09-05 | 城云科技（中国）有限公司 | Method and device for judging object based on picture area calculation and application of method and device |
| CN116704008B (en) * | 2023-08-01 | 2023-10-17 | 城云科技（中国）有限公司 | Method and device for judging object based on picture area calculation and application of method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107633526B (en) | Image tracking point acquisition method and device and storage medium | |
| CN111210429B (en) | Point cloud data partitioning method and device and obstacle detection method and device | |
| US11263445B2 (en) | Method, apparatus and system for human body tracking processing | |
| US11900676B2 (en) | Method and apparatus for detecting target in video, computing device, and storage medium | |
| CN111553946B (en) | Method and device for removing ground point cloud and method and device for detecting obstacle | |
| CN111582054A (en) | Point cloud data processing method and device and obstacle detection method and device | |
| US20180253852A1 (en) | Method and device for locating image edge in natural background | |
| CN110827292B (en) | Video instance segmentation method and device based on convolutional neural network | |
| CN113112542A (en) | Visual positioning method and device, electronic equipment and storage medium | |
| CN112052702A (en) | Method and device for identifying two-dimensional code | |
| CN114170373A (en) | Target object labeling method, processor, device and mixing station | |
| CN114708291A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| CN110223320B (en) | Object detection tracking method and detection tracking device | |
| CN114708287A (en) | Shot boundary detection method, device and storage medium | |
| CN113591827A (en) | Text image processing method and device, electronic equipment and readable storage medium | |
| CN114998768B (en) | Intelligent construction site management system and method based on unmanned aerial vehicle | |
| CN116017129A (en) | Method, device, system, equipment and medium for adjusting angle of light supplementing lamp | |
| CN112529943B (en) | Object detection method, object detection device and intelligent equipment | |
| CN112016609B (en) | Image clustering method, device, equipment and computer storage medium | |
| US20220027623A1 (en) | Object Location Determination in Frames of a Video Stream | |
| US20140050401A1 (en) | Fast Image Processing for Recognition Objectives System | |
| CN113449697B (en) | Video information processing method, device and computer storage medium | |
| CN117218162B (en) | Panoramic tracking vision control system based on ai | |
| CN115830299A (en) | Image recognition method and device, storage medium and processor | |
| CN113887348A (en) | Target detection method and device of camera array, electronic equipment and storage medium |
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 |