Intelligent labeling method, electronic equipment and storage medium

Abstract

The invention discloses an intelligent labeling method, electronic equipment and a storage medium, and the intelligent labeling method comprises the steps: carrying out the intelligent labeling of the type and shape of a first labeling object on a lane in response to the operation of a user on a labeling starting assembly on a labeling interface; in response to completion of intelligent labeling, displaying a second labeling object of which the type and / or shape cannot be intelligently labeled in the first labeling object to a labeling task list of the labeling interface; on the basis of interaction with the user, manually labeling the type and / or shape of a second labeling object in the labeling task list; and in response to completion of manual labeling, performing quality inspection on the type and the shape of the first labeled object.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of lane marking, and in particular to an intelligent marking method, electronic device, and storage medium. Background Art

[0002] In the related technology, in the process of developing intelligent driving systems, it is often necessary to provide a large amount of lane annotation data to train the system. Traditional lane annotation data is usually manually annotated by annotators, which is not only time-consuming but also requires a lot of professional knowledge and annotation skills.

[0003] In the related technology, the commonly used lane labeling method is "manually identify the location and type of the annotation object -> manually operate the labeling process to label the lane data". This labeling method has the following disadvantages: Low manual labeling efficiency: Manually labeling lanes requires a lot of time and effort, especially in complex lane scenes. The labeler needs to manually identify and manually label each annotation object, which greatly reduces work efficiency; Low quality of labeling results: During the manual labeling process, the labeler may miss the annotation object due to fatigue or lack of concentration, affecting the accuracy of the final data; High professional knowledge and skills requirements: Manual labeling requires the labeler to have certain professional knowledge and skills, including the ability to recognize lane features and the proficiency of labeling tools. For beginners or non-professionals, the labeling process has a high threshold and it is difficult to get started quickly. Summary of the invention

[0004] The embodiments of the present invention provide an intelligent labeling method, an electronic device, and a storage medium, which are used to solve at least one of the above technical problems.

[0005] In a first aspect, an embodiment of the present invention provides an intelligent labeling method, including: in response to a user's operation of a start labeling component on a labeling interface, intelligently labeling the type and shape of a first labeling object on a lane; in response to completion of intelligent labeling, displaying a second labeling object in the first labeling object whose type and / or shape cannot be intelligently labeled in a labeling task list of the labeling interface; based on interaction with the user, manually labeling the type and / or shape of the second labeling object in the labeling task list; in response to completion of manual labeling, performing quality inspection on the type and shape of the first labeling object.

[0006] In a second aspect, an embodiment of the present invention provides an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any one of the above-mentioned intelligent labeling methods of the present invention.

[0007] In a third aspect, an embodiment of the present invention provides a storage medium, in which one or more programs including execution instructions are stored. The execution instructions can be read and executed by an electronic device (including but not limited to a computer, a server, or a network device, etc.) to execute any of the above-mentioned intelligent labeling methods of the present invention.

[0008] In a fourth aspect, an embodiment of the present invention further provides a computer program product, wherein the computer program product comprises a computer program stored on a storage medium, wherein the computer program comprises program instructions, and when the program instructions are executed by a computer, the computer executes any one of the above-mentioned intelligent labeling methods.

[0009] In a fifth aspect, an embodiment of the present invention further provides a movable platform, comprising the electronic device as described in the second aspect.

[0010] The method of the present application realizes intelligent visual automatic labeling by intelligently identifying the type and location of the labeled object, providing visual labeling progress information display and automatically labeling the object. Furthermore, it can also provide a labeling task list to display objects that cannot be automatically labeled, provide a help interface to help labelers learn relevant labeling knowledge, and display labeling methods according to different object types to assist labelers in labeling, so that labelers can complete labeling more easily. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0012] Figure 1 A flowchart of an intelligent labeling method provided by one embodiment of the present invention; Figure 2a-2g A flowchart of using intelligent automatic labeling of a specific example of the intelligent labeling method provided by one embodiment of the present invention; Figure 3a-3e A flowchart of auxiliary annotation usage of a specific example of the intelligent annotation method provided by one embodiment of the present invention; Figure 4 A functional schematic diagram of a specific example of an intelligent annotation method provided by an embodiment of the present invention, in which annotation knowledge and a tool use method are provided; Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0013] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0014] Autonomous driving lane marking plays a vital role in the autonomous driving system. It uses precise image processing technology to ensure that the vehicle can accurately identify lane lines and objects and signs around the lanes, thereby ensuring driving safety and stability. The main purpose of lane marking is to provide the autonomous driving system with detailed road information so that it can make correct decisions under various road conditions. Among them, the lane marking of this application is not limited to marking lanes, but also objects around lanes (such as traffic lights, toll poles, railings, signs, etc.). In other scenarios, it may not only be objects around the lanes, but also pedestrians, vehicles and even other objects such as animals and trees can be marked, and this application is not limited here.

[0015] In order to solve one or more problems existing in manual labeling, the embodiment of the present application proposes a function of intelligent automatic lane labeling and auxiliary labeling. By externalizing the process of intelligent automatic labeling and the task list of auxiliary identification of labeled objects, it helps the labeler to complete the labeling quickly and reduces the dependence on the professional skills of the labeler.

[0016] Please refer to Figure 1 , which shows a flow chart of an intelligent labeling method provided by an embodiment of the present invention. The intelligent labeling method may exist in a computer in the form of an algorithm, and in subsequent embodiments, it may be referred to as an intelligent labeling device, such as a program code or a software, etc., and the present application is not limited thereto.

[0017] like Figure 1 As shown, in step 101, in response to the user's operation on the start annotation component on the annotation interface, the type and shape of the first annotation object on the lane are intelligently annotated; In step 102, in response to completion of intelligent annotation, a second annotation object of the first annotation object whose type and / or shape cannot be intelligently annotated is displayed in the annotation task list of the annotation interface; In step 103, based on the interaction with the user, manually annotate the type and / or shape of the second annotation object in the annotation task list; In step 104, in response to the completion of manual annotation, a quality inspection is performed on the type and shape of the first annotation object.

[0018] In this embodiment, for step 101, the intelligent labeling device responds to the user's operation on the start labeling component on the labeling interface, and intelligently labels the type and shape of the first labeling object. Among them, the first labeling object may include a variety of elements on the lane, such as lane groups, lane lines, lanes (including intersections), intersection topological relationships, road edges, road surface areas, road surface markings, drivable / non-drivable areas, intersection areas, invalid areas, parking spaces, parking space markings, 3D bounding boxes, poles / signs, traffic lights, speed limits, etc., and may also include objects around the lane, such as traffic lights, toll poles, railings, signboards, etc. In other scenarios, it may be possible to mark more than just objects around the lane, and pedestrians, vehicles, and even other objects such as animals and trees, which are not limited in this application. For example, when the user clicks the start labeling button, the system starts intelligent automatic labeling, and can also display the progress of automatic labeling, so that the user can promptly understand the type of labeling object currently automatically labeled and the number of completed labels. In a specific example, after entering the labeling task, it automatically switches to the automatic labeling assistant and displays the automatic labeling function. Then, in response to the user's operation on the start labeling component on the labeling interface, automatic labeling starts and displays the labeling progress. After the labeling is completed, the automatic labeling result page is displayed, and the screen synchronously displays all automatically marked objects.

[0019] Then, for step 102, the intelligent labeling device responds to the completion of intelligent labeling by displaying the second labeling object whose type and / or shape cannot be intelligently labeled in the first labeling object to the labeling task list of the labeling interface. For example, the intersection area and the road edge line cannot be automatically labeled, and the intersection area and the road edge line are displayed in the labeling task list of the labeling interface. Among them, the reason why there are objects that cannot be intelligently labeled is mainly because the existing algorithm cannot recognize all the labeling objects. Therefore, there may be a situation where the shape of the object cannot be accurately identified and the shape cannot be automatically drawn, but it can be detected that there is an object that needs to be labeled at the corresponding position. After the algorithm is improved in the future, if all the labeling objects can be identified, this step can be omitted or it can be displayed that there is no second labeling object. This application is not limited here.

[0020] Afterwards, for step 103, the intelligent annotation device manually annotates the type and / or shape of the second annotation object in the annotation task list based on the interaction with the user. Specifically, there may be shapes that cannot be recognized, there may be types that cannot be recognized, and there may be shapes and types that cannot be intelligently recognized. This application is not limited here. In a specific example, the user uses the auxiliary annotation function and clicks on the annotation task list. The system will position the annotation objects in the middle of the screen in turn, and the annotator can manually annotate them in turn, without having to search for the annotation objects one by one in the large screen.

[0021] Finally, for step 104, the intelligent labeling device performs quality inspection on the type and shape of the first labeled object in response to the completion of manual labeling. In a specific example, after the labeling is completed, a prompt for quality inspection attention requirements is displayed. When the user clicks on automatic quality inspection, the interface switches to the question tab to display a list of problems automatically inspected. For example, the shapes, attributes, directions of lines, and relationships between objects of various types all have rules and correct references at the technical level. When using automatic quality inspection, the system will detect the "labeled objects" with the "correct reference rules". If there is a mismatch, the wrongly labeled object will be displayed in the problem list as "object name". In addition, the labeling data for the above-mentioned second labeled object or the labeling data for the second labeled object that is correctly labeled after quality inspection can be stored and output as training data to train and update the intelligent labeling algorithm, and the updated intelligent labeling algorithm can be imported into the intelligent labeling device for application, which can improve the labeling efficiency.

[0022] The method of this embodiment improves the labeling efficiency and quality by intelligently identifying the type and shape of the labeled object, and displaying the objects that cannot be labeled automatically, completing the labeling of the objects that cannot be labeled automatically through interaction with the user, and then performing quality inspection on the labeled objects.

[0023] In some optional embodiments, in response to the completion of intelligent annotation, the method further includes: locating and displaying the first first annotation object to be checked on the annotation interface; and based on the interaction with the user, checking and / or modifying the type and shape of the first annotation object of the intelligent annotation in turn.

[0024] In one embodiment, after the automatic annotation is completed, the result page of the automatic annotation can be displayed, and the screen can synchronously display all the automatically marked objects, automatically select the automatically labeled objects to be checked, and display the selected objects in the center, and then check and / or modify them by the user. After the inspection and / or modification is completed, the user's instructions to check the next one are accepted until the inspection and / or modification of all objects are completed. In another embodiment, when displaying the second annotated object, if the user's inspection instruction is received, for example, the inspection button is clicked, the result page of the automatic annotation is displayed at this time, and the first first annotated object to be checked is located and displayed on the annotation interface until the inspection and / or modification of all objects is completed through interaction with the user. It can be entered into the manual inspection and modification link automatically, or it can be entered at any time through the user's operation, and the present application is not limited here.

[0025] It should be noted that the order of execution of both manual labeling and quality inspection is not fixed. It can be automatically connected in certain links or can be entered at any time according to user needs. This application does not limit this. For example, regarding the inspection of automatic labeling and manual labeling, you can first check the manually labeled objects, and then check the automatically labeled objects; you can also check the automatically labeled and manually labeled objects together; you can also only check the automatically labeled objects or only check the manually labeled objects. This application does not limit this.

[0026] The method of the present application locates and displays the annotation objects to be checked on the annotation interface, and checks and / or modifies the types and shapes of the annotation objects in turn based on the interaction with the user, thereby greatly reducing the need for manual annotation.

[0027] In some optional embodiments, during manual labeling and / or inspection, the method further includes: the intelligent labeling device displays the current target in the center, wherein the target is the second labeling object in the labeling task list and / or the first labeling object of the intelligent labeling; the intelligent labeling device displays the next target in the center in response to the completion of the current target processing, until all target processing is completed. In the manual labeling process, the target is the second labeling object in the labeling task list, and in the inspection process, the target is the first labeling object in the labeling task list that has completed intelligent labeling. For example, after identifying the object to be labeled, the xy axis position value of each object in the page will be recorded in the background. When a task in the labeling task list is clicked, the xy axis position value of the object to be labeled in the task will be automatically identified and displayed in the middle of the screen; click the "locate to next" button, and the xy axis position value of the next object will be obtained in a predetermined order and located in the middle of the screen. The method of this embodiment can quickly locate the object position by displaying the current target in the center.

[0028] In some optional embodiments, after displaying the second annotation object whose type and / or shape cannot be intelligently annotated in the first annotation object to the annotation task list of the annotation interface, the method further includes: the intelligent annotation device locates and displays any annotation task in response to the user's operation on any annotation task in the annotation task list. For example, when the user clicks on any annotation task in the annotation task list, the system will locate the annotation object corresponding to the annotation task in the middle of the screen in turn, and the annotator can manually annotate in turn, without having to search for the annotation object one by one in the large screen. Alternatively, the corresponding position can be highlighted (or boxed) in a large picture, and the task information can be displayed. After the user selects the corresponding position, the area to be annotated can be directly displayed (or directly displayed without clicking), and the present application is not limited here. Among them, the annotation task list can include the following content: the total number of objects to be annotated (for example: 43 objects to be annotated), the type of object to be annotated and the number of objects of this type to be annotated (for example: road edge line (30)), the annotation method of the object type, displayed in an icon form, and can also include other content, and the present application is not limited here. This can help the labeler to check the labeling results one by one by clicking buttons in an independent window, greatly reducing the difficulty of manual labeling.

[0029] In some optional embodiments, quality inspection of the type and shape of the first annotation object includes: the intelligent annotation device determines the correct reference object rule corresponding to the first annotation object based on the type of the first annotation object, wherein the correct reference object rule includes the shape of the first annotation object, the attribute of the first annotation object, the direction of the line of the first annotation object and / or the relationship between the first annotation object and other annotation objects; then, the intelligent annotation device matches the first annotation object completed by intelligent annotation and / or manual annotation with the corresponding correct reference object rule; if the first annotation object does not match the corresponding correct reference object rule, the unmatched first annotation object is displayed. In addition to displaying the unmatched first annotation object, an error list can also be displayed. When the annotator chooses to view the corresponding error task, the corresponding position can be located and the correct reference object rule can be given for the annotator to modify. The annotator can directly modify the annotation of the unmatched first annotation object, thereby reducing the learning cost of the annotator and improving the annotation efficiency.

[0030] In some optional embodiments, the annotation interface also includes a first help component, and the method also includes: the intelligent annotation device responds to the user's operation on the first help component, and displays the annotation knowledge and / or the usage of each annotation tool for the user to view. For example, when the annotator clicks the "Help" button on the page, it will jump to the help center to view the annotation professional knowledge and tool usage tutorials (such as documents, videos, etc.), so as to help the annotator recall how to annotate a certain type of object in time, and also lower the annotation threshold for beginners or non-professionals, so that they can get started faster and improve annotation efficiency.

[0031] In some optional embodiments, each annotated object in the annotation task list has a corresponding second help component, and the method further includes: in response to the user's operation on the second help component, displaying the annotation method of the annotation object type associated with the annotation object corresponding to the second help component. For example, the annotator clicks on the annotation object type in the task list to view the annotation method of the object separately, so as to help the annotator recall how to annotate a certain type of object in time, and also lower the annotation threshold for beginners or non-professionals, so that they can get started faster and improve the annotation efficiency.

[0032] The embodiment of the present application proposes a visual display method for intelligent automatic lane marking. On the automatic marking sidebar interface, the automatic marking progress and results are displayed in the form of graphics or text, and the user is allowed to click a button in an independent window to check the marking results one by one. The embodiment of the present application also proposes an interactive use process for automatic lane marking. The user clicks a button and the system will automatically mark. Then, the user checks the marking results in an independent window. In the whole process, the demand for manual marking is greatly reduced, and the marking efficiency is effectively improved. The embodiment of the present application also proposes an interactive use process for auxiliary marking. The system automatically identifies and displays a list of objects to be marked in the sidebar. When the user clicks the list, the system will display the object in the center. The user can mark manually immediately, reducing the time of manually finding the location of the object. When the marking is completed, the user clicks on the automatic quality inspection, and the system will automatically check the problems of manual marking. The embodiment of the present application also proposes a function of viewing the marking method tutorial in real time in the form of video. The user can click a button at any time during the marking process to view the marking method tutorial in real time on the current page, which can help the labeler recall how to mark a certain type of object in time, and also facilitate beginners or non-professionals to get started faster, which helps to improve the marking efficiency.

[0033] The following specific interfaces are given to illustrate the embodiments of the present application so that those skilled in the art can better understand the solution of the present application. It should be noted that the following interfaces and related descriptions are not used to limit the scope of the present application. Those skilled in the art can understand that the solution of the present application can also be implemented using other similar interfaces and methods.

[0034] Please refer to Figure 2a-2g , which shows a usage flow chart of intelligent automatic labeling of a specific example of an intelligent lane labeling method provided by an embodiment of the present invention.

[0035] First, Figure 2a As shown, enter the annotation task and automatically switch to the AI ​​annotation assistant. The large area on the left is the viewing area, where the annotation object is located, and the small area on the right is used to display various operation interfaces and parameters. Then, Figure 2b The AI ​​automatic annotation function is demonstrated. Furthermore, the annotation progress can be displayed. Figure 2b After clicking the "Start AI Labeling" button in Figure 2c The progress of annotation is displayed in Figure 2d As shown in the figure, when the labeling is completed, the automatic labeling result page is displayed, and all the automatically labeled objects are displayed synchronously on the screen. Later, when checking the automatically labeled objects, when the user clicks Figure 2d Click the "Go to Check" button in Figure 2e The object to be inspected is automatically selected and displayed in the center. Figure 2e After clicking the "Next" button in Figure 2f The middle viewing area moves smoothly in the direction of the "driving track" to center the next object to be inspected. Figure 2g As shown, until all objects are checked, the completion prompt will be displayed in the independent window in the lower right corner.

[0036] In a specific example, the annotator clicks a button, and the intelligent annotation device (system) starts intelligent automatic annotation, while displaying the progress of automatic annotation. The annotator can promptly understand the type of annotation objects currently automatically annotated and the number of annotations completed. When the system completes automatic annotation, the annotator only needs to click on the check annotation task list to check the annotation results of each annotation object. This function and usage process greatly reduces the time for manual annotation, and only the results need to be checked. Furthermore, when the annotator uses intelligent automatic annotation, the system algorithm can identify the annotation object type and automatically annotate the shape of the object, without missing objects that need to be labeled.

[0037] Please refer to Figure 3a-3e , which shows a flowchart of auxiliary labeling usage of a specific example of the intelligent lane labeling method provided by an embodiment of the present invention.

[0038] In the auxiliary annotation process, first Figure 3a As shown, the system identifies the object type and location (which cannot be automatically labeled) and displays it in the labeling task list (such as Figure 3a ), when the user clicks Figure 3a Click the "Go to next" button above the labeling task list in Figure 3bThe next object is displayed in the center, and the user manually annotates it. After all the annotations are completed, the results are displayed. Figure 3b Click the "View Quality Inspection Requirements" button in Figure 3c The quality inspection requirements are displayed in the prompt. Click Figure 3c Go to "Go to automatic quality inspection" in the Figure 3d In the quality inspection process interface, after the quality inspection is completed, Figure 3e As shown, the interface in the lower right corner switches to the Problems tab, showing a list of problems found by automatic quality inspection.

[0039] Automatic quality inspection algorithm implementation process: The shapes, attributes, directions of lines, and relationships between various objects have rules and correct references at the technical level; when using automatic quality inspection, the system will check the "labeled objects" with the "correct reference rules". If there is a mismatch, the incorrectly labeled object will be displayed in the problem list as "object name" (example object name: b467897676a).

[0040] Among them, the intelligent labeling device will automatically identify the tasks that should be labeled and automatically display them in the form of a list, namely the "labeling task list". Specifically, the labeling task list may include the following: the total number of objects to be labeled (for example: 43 objects to be labeled), the type of objects to be labeled and the number of objects of this type to be labeled (for example: road edge lines (30)), the labeling method of the object type, displayed in the form of an icon, and may also include other content, which is not limited in this application. In this example, the objects that cannot be automatically labeled temporarily include intersection areas and road edge lines. The reason for the inability to automatically label is: the algorithm recognition accuracy is less than 85%, so it is impossible to accurately identify the shape of the object and automatically draw the shape. In the future, the algorithm recognition accuracy will be higher, and the number of objects that cannot be automatically identified will also decrease, or there will no longer be objects that cannot be automatically identified. In this case, manual inspection is only required, and there is no need for auxiliary labeling.

[0041] In a specific example, the annotator uses the auxiliary annotation function and clicks on the annotation task list. The system will position the annotation objects in the middle of the screen one by one, and the annotator can manually annotate them one by one, without having to search for the annotation objects one by one in the large screen. For example, the system automatically identifies and displays the list of objects to be annotated in the sidebar. When the user clicks on the list, the system will center the object, and the user can manually annotate immediately, reducing the time of manually searching for the location of the object; when the annotation is completed, the user clicks on automatic quality inspection, and the system will automatically check the problems of manual annotation.

[0042] Please refer to Figure 4 , which shows a functional schematic diagram of annotation knowledge and tool usage method of a specific example of an intelligent annotation method provided by an embodiment of the present invention.

[0043] When the user clicks Figure 3a Click the "Help" icon in the upper right corner to view annotation knowledge and tool usage tutorials, which include annotation process specifications for various annotation tools or annotators. Figure 3a When you click the "Object Type Annotation Method" button (a small button next to each annotation task without any text) in the annotation task list, a pop-up window displays multimedia files related to the annotation method of the object type, such as Figure 4 Videos in .

[0044] In a specific example, when the annotator clicks the "Help" button on the page, he will be redirected to the help center to view the annotation professional knowledge and tool usage tutorials. In addition, the annotator can click the annotation object type in the task list to view the annotation method of the object separately. This function can help the annotator recall how to annotate a certain type of object in time; it also lowers the annotation threshold for beginners or non-professionals, allowing them to get started faster.

[0045] In the labeling method of the related art, the labeler needs to manually identify and label each annotation object in the lane. This process is time-consuming and requires high professional skills of the user. With the method of the present invention, the labeler uses a simple interface and the system will automatically complete the labeling. In addition, the present invention also provides an auxiliary labeling function. When the labeler clicks on the list of objects to be labeled, the system will quickly locate the object and display the labeling method in text or video form. The invention improves the labeling efficiency, improves the labeling quality to avoid missing labeled objects, and reduces the learning cost of the labeler.

[0046] In other embodiments, the present invention also provides a non-volatile computer storage medium, the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the intelligent labeling method in any of the above method embodiments; As an implementation mode, the non-volatile computer storage medium of the present invention stores computer executable instructions, and the computer executable instructions are configured as follows: In response to the user's operation of the start annotation component on the annotation interface, intelligently annotate the type and shape of the first annotation object on the lane; In response to the completion of intelligent annotation, displaying a second annotation object of the first annotation object whose type and / or shape cannot be intelligently annotated in the annotation task list of the annotation interface; Based on the interaction with the user, manually annotating the type and / or shape of the second annotation object in the annotation task list; In response to the manual annotation being completed, a quality inspection is performed on the type and shape of the first annotation object.

[0047] The non-volatile computer-readable storage medium may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the intelligent labeling device, etc. In addition, the non-volatile computer-readable storage medium may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the non-volatile computer-readable storage medium may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the intelligent labeling device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0048] An embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes any one of the above-mentioned intelligent labeling methods.

[0049] Figure 5 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 5 As shown, the device includes: one or more processors 510 and a memory 520, Figure 5 A processor 510 is used as an example. The device of the intelligent labeling method may also include: an input device 530 and an output device 540. The processor 510, the memory 520, the input device 530 and the output device 540 may be connected via a bus or other means. Figure 5 In the example, the connection through the bus is taken. The memory 520 is the above-mentioned non-volatile computer-readable storage medium. The processor 510 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 520, that is, the intelligent labeling method of the above-mentioned method embodiment is realized. The input device 530 can receive input digital or character information, and generate key signal input related to user settings and function control of the intelligent labeling device. The output device 540 may include a display device such as a display screen.

[0050] The above product can execute the method provided by the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided by the embodiment of the present invention.

[0051] As an implementation mode, the electronic device is applied to a smart labeling device, comprising: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: In response to the user's operation of the start annotation component on the annotation interface, intelligently annotate the type and shape of the first annotation object on the lane; In response to the completion of intelligent annotation, displaying a second annotation object of the first annotation object whose type and / or shape cannot be intelligently annotated in the annotation task list of the annotation interface; Based on the interaction with the user, manually annotating the type and / or shape of the second annotation object in the annotation task list; In response to the manual annotation being completed, a quality inspection is performed on the type and shape of the first annotation object.

[0052] The electronic device of the embodiment of the present application exists in various forms, including but not limited to: (1) Mobile communication devices: These devices are characterized by their mobile communication functions and their main purpose is to provide voice and data communications. These terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones.

[0053] (2) Ultra-mobile personal computer devices: These devices fall into the category of personal computers, have computing and processing capabilities, and generally also have mobile Internet access features. These terminals include: PDA, MID and UMPC devices, such as iPad.

[0054] (3) Portable entertainment devices: These devices can display and play multimedia content. They include audio and video players (such as iPods), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.

[0055] (4) Server: A device that provides computing services. The server consists of a processor, hard disk, memory, system bus, etc. The server is similar to a general computer architecture, but because it needs to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.

[0056] (5) Other electronic devices with data interaction functions.

[0057] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0058] Through the description of the above implementation modes, those skilled in the art can clearly understand that each implementation mode can be implemented by means of software plus a necessary general hardware platform, or of course by hardware. Based on such an understanding, the above technical solution can essentially or in other words be embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent labeling method, comprising: In response to the user's operation of starting the annotation component on the annotation interface, intelligently annotating the type and shape of the first annotation object; In response to the completion of intelligent annotation, displaying a second annotation object of the first annotation object whose type and / or shape cannot be intelligently annotated in the annotation task list of the annotation interface; Based on the interaction with the user, manually annotating the type and / or shape of the second annotation object in the annotation task list; In response to the manual annotation being completed, a quality inspection is performed on the type and shape of the first annotation object.

2. The method according to claim 1, wherein: In response to the intelligent annotation being completed, the method further includes: Positioning and displaying a first first annotation object to be checked on the annotation interface; Based on the interaction with the user, the type and shape of the first annotation object of the intelligent annotation are checked and / or modified in sequence.

3. The method according to claim 2, wherein: During the manual annotation and / or inspection process, the method further includes: Centering the current target, wherein the target is the second annotation object in the annotation task list and / or the first annotation object of the intelligent annotation; In response to the current target being processed, the next target is displayed in the center until all targets are processed.

4. The method according to claim 1, wherein: After displaying the second annotation object of the first annotation object whose type and / or shape cannot be intelligently annotated in the annotation task list of the annotation interface, the method further includes: In response to the user's operation on any labeling task in the labeling task list, the any labeling task is located and displayed.

5. The method according to claim 1, wherein: The quality inspection of the type and shape of the first annotation object includes: Determining a correct reference object rule corresponding to the first annotation based on the type of the first annotation, wherein the correct reference object rule includes a shape of the first annotation, an attribute of the first annotation, a direction in which a line of the first annotation points, and / or a relationship between the first annotation and other annotations; Matching the first annotated object that has been intelligently annotated and / or manually annotated with a corresponding correct reference object rule; If the first annotation does not match the corresponding correct reference rule, the unmatched first annotation is displayed.

6. The method according to any one of claims 1 to 5, wherein: The annotation interface also includes a first help component, and the method further includes: In response to the user's operation on the first help component, annotation knowledge and / or methods of using each annotation tool are displayed for the user to view.

7. The method according to any one of claims 1 to 5, wherein: Each annotation object in the annotation task list has a corresponding second help component, and the method further includes: In response to the user's operation on the second help component, a marking method of a marking object type associated with the marking object corresponding to the second help component is displayed.

8. An electronic device, comprising: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the steps of the method described in any one of claims 1 to 7.

9. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A movable platform comprising the electronic device as described in claim 8.

Images