阅读说明：本技术 应用于智慧物流领域的数据标注方法及平台 (Data annotation method and platform applied to intelligent logistics field ) 是由 余琴 于 2021-07-26 设计创作，主要内容包括：本发明涉及一种应用于智慧物流领域的数据标注方法及平台,属于数据标注技术领域,该方法通过获取物流图片；根据预设间隔,在物流图片中抽取有效图片；通过人工识别对有效图片进行分类,以使标注人员对每类别图片的类别进行标注；生成标注任务,将标注任务绑定预设的多个标注人员；接收标注人员的标注信息,对有效图片进行标注。本申请实现对图片的智能化标注和辅助标注,在线生成任务,把人与任务相关联,实现对图片的便捷标注。(The invention relates to a data annotation method and a platform applied to the field of intelligent logistics, belonging to the technical field of data annotation, wherein the method comprises the steps of obtaining a logistics picture; extracting effective pictures from the logistics pictures according to a preset interval; classifying the effective pictures through manual identification so that a labeling person labels the class of each class of picture; generating a labeling task, and binding the labeling task to a plurality of preset labeling personnel; and receiving the marking information of the marking personnel, and marking the effective picture. The method and the device realize intelligent marking and auxiliary marking of the picture, generate the task on line, associate people with the task, and realize convenient marking of the picture.)

1. A data labeling method applied to the field of intelligent logistics is characterized by comprising the following steps:

acquiring a logistics picture;

extracting effective pictures from the logistics pictures according to a preset interval;

classifying the effective pictures through manual identification so that a labeling person labels the classes of the pictures of each class;

generating a labeling task, and binding the labeling task with a plurality of preset labeling personnel;

and receiving the marking information of the marking personnel, and marking the effective picture.

2. The method of claim 1, further comprising, after said obtaining the logistics picture:

judging whether the logistics picture contains preset content or not, wherein the preset content comprises: at least one of a person, a package, a vehicle;

if the preset content is not contained in the logistics picture, determining that the logistics picture is an invalid picture;

and deleting the invalid picture.

3. The method according to claim 1, wherein the receiving labeling information of the labeling person and labeling the effective picture comprises:

receiving a labeling instruction of the labeling personnel, and displaying different labels in different colors through label customization;

and displaying the marked effective picture at a preset position so that the marking personnel revises the marked effective picture.

4. The method according to claim 3, wherein the receiving labeling information of the labeling personnel and labeling the effective picture comprises:

marking similar objects with the same color through a preset algorithm, and checking the marked objects through manual checking; and the number of the first and second groups,

and marking the things in the effective picture through semi-intelligent tool identification.

5. The method according to claim 1, wherein after receiving the labeling information of the labeling person and labeling the effective picture, the method further comprises:

generating an auditing task according to the marked effective picture so that a corresponding auditor audits the marked effective picture;

and generating an audit result report after the marked effective picture is audited or sampling audit is completed.

6. The method of claim 5, further comprising:

and performing model training according to the labeling result of the labeled effective picture to obtain a labeled model.

7. The utility model provides a be applied to data annotation platform in wisdom commodity circulation field, includes: a processor, and a memory coupled to the processor;

the memory is used for storing a computer program, and the computer program is at least used for executing the data annotation method applied to the intelligent logistics field in any one of claims 1 to 6;

the processor is used to call and execute the computer program in the memory.

Technical Field

The invention belongs to the technical field of data annotation, and particularly relates to a data annotation method and a data annotation platform applied to the field of intelligent logistics.

Background

With the development of science and technology, data annotation becomes an important link of machine learning. For example, for the recognition of an image by machine vision, certain labeled sample data is needed, in short, a certain amount of textbook is needed, the algorithm forms own knowledge ability by learning the textbook (labeled sample data), and when a relevant scene is met, judgment is automatically made. The marking of the sample data is carried out manually at present, information on the picture is classified and marked manually, and marked content is transmitted to a machine vision algorithm in a data form to serve as a standard of machine vision learning. Therefore, the labeling quality of the sample data directly affects the recognition effect of machine vision, and is very important.

However, most of the existing open-source labeling tools in the market are deployed in foreign servers, so that the open mode is difficult, and the data security problem is also involved. Therefore, how to label data quickly and safely becomes a technical problem to be solved urgently in the prior art.

Disclosure of Invention

The invention provides a data labeling method and a data labeling platform applied to the field of intelligent logistics, and aims to solve the technical problems of difficult labeling and low labeling safety in the prior art.

The technical scheme provided by the invention is as follows:

on one hand, the data annotation method applied to the field of intelligent logistics comprises the following steps:

acquiring a logistics picture;

extracting effective pictures from the logistics pictures according to a preset interval;

classifying the effective pictures through manual identification so that a labeling person labels the classes of the pictures of each class;

generating a labeling task, and binding the labeling task with a plurality of preset labeling personnel;

and receiving the marking information of the marking personnel, and marking the effective picture.

Optionally, after the acquiring the logistics image, the method further includes:

judging whether the logistics picture contains preset content or not, wherein the preset content comprises: at least one of a person, a package, a vehicle;

if the preset content is not contained in the logistics picture, determining that the logistics picture is an invalid picture;

and deleting the invalid picture.

Optionally, the receiving the labeling information of the labeling personnel and labeling the effective picture includes:

receiving a labeling instruction of the labeling personnel, and displaying different labels in different colors through label customization;

and displaying the marked effective picture at a preset position so that the marking personnel revises the marked effective picture.

Optionally, the receiving the labeling information of the labeling personnel and labeling the effective picture includes:

marking similar objects with the same color through a preset algorithm, and checking the marked objects through manual checking; and the number of the first and second groups,

and marking the things in the effective picture through semi-intelligent tool identification.

Optionally, the receiving the labeling information of the labeling personnel, labeling the effective picture, and then, further including:

generating an auditing task according to the marked effective picture so that a corresponding auditor audits the marked effective picture;

and generating an audit result report after the marked effective picture is audited or sampling audit is completed.

Optionally, the method further includes:

and performing model training according to the labeling result of the labeled effective picture to obtain a labeled model.

In another aspect, a data annotation platform applied to the field of intelligent logistics includes: a processor, and a memory coupled to the processor;

the memory is used for storing a computer program, and the computer program is at least used for executing any one of the data labeling methods applied to the intelligent logistics field;

the processor is used to call and execute the computer program in the memory.

The invention has the beneficial effects that:

the embodiment of the invention provides a data annotation method and a data annotation platform applied to the field of intelligent logistics, wherein the method comprises the steps of obtaining a logistics picture; extracting effective pictures from the logistics pictures according to a preset interval; classifying the effective pictures through manual identification so that a labeling person labels the class of each class of picture; generating a labeling task, and binding the labeling task to a plurality of preset labeling personnel; and receiving the marking information of the marking personnel, and marking the effective picture. The method and the device realize intelligent marking and auxiliary marking of the picture, generate the task on line, associate people with the task, and realize convenient marking of the picture.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flowchart of a data annotation method applied in the field of intelligent logistics according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a data annotation platform applied to the field of intelligent logistics according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without any inventive step, are within the scope of the present invention.

In order to at least solve the technical problem provided by the invention, the embodiment of the invention provides a data annotation method applied to the field of intelligent logistics.

Fig. 1 is a schematic flow chart of a data annotation method applied in the field of intelligent logistics according to an embodiment of the present invention, as shown in fig. 1, the method according to the embodiment of the present invention may include the following steps:

and S11, acquiring logistics pictures.

For example, the mode of acquiring the logistics picture can be manual introduction, and can also be transmitted through a secure encryption interface, so that the convenience and the safety are realized.

And S12, extracting effective pictures from the logistics pictures according to the preset intervals.

For example, after the logistics picture is acquired, the logistics picture can be subjected to data cleaning. For example, the target data is intelligently identified through an intelligent algorithm of a labeling platform, invalid data is removed, for example, data without vehicles in the picture data of the vehicle loading rate can be removed through a vehicle identification algorithm, time is saved, and labeling efficiency is improved.

In some embodiments, optionally, after the obtaining of the logistics picture, the method further includes: judging whether the logistics picture contains preset content or not, wherein the preset content comprises: at least one of a person, a package, a vehicle; if the logistics picture does not contain the preset content, determining the logistics picture as an invalid picture; and deleting the invalid picture.

In the application, data is cleaned by deleting invalid pictures. After the data is cleaned, effective pictures can be extracted and marked according to a preset interval. The preset interval may be a specified duration or an unfixed duration, and is not specifically limited herein.

And S13, classifying the effective pictures through manual identification so that the labeling personnel label the category of each category of pictures.

For example, after sample data is provided, manual identification is often preferentially performed, for example, a picture of vehicle identification needs to be subjected to vehicle marking, a sample picture of a license plate needs to be subjected to license plate marking, and one or a batch of sample data also needs to be subjected to multiple marking.

And S14, generating a labeling task, and binding the labeling task to a plurality of preset labeling personnel.

For example, people can be bound with annotation tasks, the platform automatically distributes the tasks to a plurality of annotators, and the manager background can check member progress and find errors in real time.

And S15, receiving the labeling information of the labeling personnel, and labeling the effective pictures.

In some embodiments, optionally, receiving annotation information of an annotation person, and annotating the effective picture, includes:

receiving a labeling instruction of a labeling person, and displaying different labels in different colors through label customization;

and displaying the marked effective picture at a preset position so that a marking person can revise the marked effective picture.

In some embodiments, optionally, receiving annotation information of an annotation person, and annotating the effective picture, includes:

marking similar objects with the same color through a preset algorithm, and checking the marked objects through manual checking; and the number of the first and second groups,

and marking the things in the effective picture through semi-intelligent tool identification.

For example, a annotator can annotate things in the active picture by: the marking personnel can click the frame selection object by using a mouse, and the shortcut key operation is supported. The label is self-defined, different labels are distinguished through different colors, and the visual differentiation is obvious. The time interval can be selected, so that the invalid picture extracted on line is avoided, and the labeled sample with the maximum efficiency is realized. And the left side of the labeling result is displayed, so that modification and error correction are facilitated. In the marking link, an automatic identification marking tool and an auxiliary marking tool are provided besides the pure manual marking. The automatic identification marking can be realized by means of a mature algorithm before identification when marking objects with the same color and the same similarity, and only manual verification is needed, for example, a Zhongtong personnel identification algorithm is mature, when sample data of personnel needs to be marked smoothly, automatic identification can be realized by using the existing algorithm, and the identification result needs manual verification in view of the situation that a scene is obstructed. Supplementary marking tool, the instrument discernment of semi-intellectuality promptly, like square frame, polygon, triangle-shaped, straight line, curve, broken line, 3D, continuous frame mark etc. the mark of garden road can direct the curve and carry out semi-automatization's discernment, and the manual work is simply adjusted can.

In some embodiments, optionally, after receiving the annotation information of the annotating person and annotating the effective picture, the method further includes:

generating an auditing task according to the marked effective picture so that a corresponding auditor audits the marked effective picture;

and after the marked effective picture is audited or sampling audited, generating an audit result report.

For example, data annotation results require manual review. And the platform automatically generates an audit task, detects that all pictures are confirmed or audited in a sampling mode, and automatically generates a report according to the audit result.

Optionally, the method further includes: and performing model training according to the labeling result of the labeled effective picture to obtain a labeled model.

For example, after data samples are derived, a data engineer performs model training on labeled data, an algorithm analyzes each time data to obtain classification features, and a feature set very close to reality can be obtained after a large number of training.

In the present application, effect evaluation and sample optimization iteration may also be performed: the labeled data can output the result of the algorithm after model training of the algorithm, the quality of the result can be manually checked, if the quality is to be optimized, the labeling of the sample data can be increased according to the batch, and the label can be re-labeled from the data with the error identification or adjusted to be used as the next model data of the algorithm, and one optimization iteration is carried out until the standard is reached.

In the present application, a visual report may also be provided: the visual report presents historical, ongoing and non-ongoing data in the form of the report, and can carry out intelligent reminding and progress early warning in advance, evaluate the effect, the efficiency and the optimization suggestion after the fact according to the evaluation of personnel and workload in advance.

When the method is applied to the field of logistics stations, blue goods, green personnel and the like can be marked, and the method is not particularly limited. Model after the machine degree of depth learning training is finally used to the cloud aerial view platform (wisdom logistics management platform), and the scheduling and the management work of people, goods, objects such as car that appear in the video of real-time detection camera collection, the personnel vehicle of the commodity circulation transfer station of being convenient for etc.

In the present application, AILabel can be applied and extended. The AILabel is a JavaScript graphic library and supports the zooming and translation of the picture; and displaying the vector data, the text and the label on the picture.

The method comprises the steps of generating a task on line, and associating a person with the task; randomly extracting labeled pictures, avoiding invalid pictures and labeling samples with the maximum efficiency; the marked samples are visualized, error modification is supported, and the samples are more accurate; the data export format is free, and different requirements are met; intelligent labeling and auxiliary labeling; intelligent monitoring and effect evaluation; go deep into the scene of the logistics industry and refine the business process.

The embodiment of the invention provides a data annotation method applied to the field of intelligent logistics, which comprises the steps of obtaining a logistics picture; extracting effective pictures from the logistics pictures according to a preset interval; classifying the effective pictures through manual identification so that a labeling person labels the class of each class of picture; generating a labeling task, and binding the labeling task to a plurality of preset labeling personnel; and receiving the marking information of the marking personnel, and marking the effective picture. The method and the device realize intelligent marking and auxiliary marking of the picture, generate the task on line, associate people with the task, and realize convenient marking of the picture.

Based on a general inventive concept, the embodiment of the invention also provides a data annotation platform applied to the field of intelligent logistics.

Fig. 2 is a schematic structural diagram of a data annotation platform applied to the intelligent logistics field according to an embodiment of the present invention, and referring to fig. 2, the data annotation platform applied to the intelligent logistics field according to the embodiment of the present invention includes: a processor 21 and a memory 22 connected to the processor.

The memory 22 is used for storing a computer program, and the computer program is at least used for the data annotation method applied to the intelligent logistics field described in any of the above embodiments;

the processor 21 is used to invoke and execute computer programs in the memory.

The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

It is understood that the same or similar parts in the above embodiments may be mutually referred to, and the same or similar parts in other embodiments may be referred to for the content which is not described in detail in some embodiments.

It should be noted that the terms "first," "second," and the like in the description of the present invention are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Further, in the description of the present invention, the meaning of "a plurality" means at least two unless otherwise specified.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.

It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.

It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.

In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.

The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.