Label recognition method, recognition device, robot, and storage medium

Abstract

This invention provides a label identification method, wherein each label has multiple circular reflective label points. The identification method includes: acquiring a label image, the label image including one or more labels; obtaining contour lines of the label image; identifying the label points based on the contour lines; and clustering the label points to obtain the label based on the clustering results. By employing the technical solution of this invention, label points are identified through the contour lines of a label image, thereby identifying the label. Compared with existing technologies, this method is less affected by lighting conditions, yields more accurate identification results, and requires less computation. This improves the efficiency of identification devices and robots using this method, ultimately enhancing the customer experience.

Description

Technical Field

[0001] This invention relates generally to the field of image processing technology, and more particularly to a label recognition method, a label recognition device, a robot, and a computer-readable storage medium. Background Technology

[0002] With the rapid development of artificial intelligence technology, robots have been widely used in various scenarios, bringing convenience to people's work and life. Among them, positioning technology is a key technology for robots. Currently, many robots use tag recognition for positioning; therefore, the ability to accurately identify tags is crucial for robots.

[0003] Existing label recognition typically employs traditional image processing methods (such as smoothing filtering, threshold calculation, binarization, etc.). This involves filtering out scattered noise, calculating a threshold for the grayscale histogram, performing binarization based on the threshold, and then applying nearest-neighbor clustering to group light spots into labels. However, this recognition method is significantly affected by lighting conditions and the label points themselves, and it is computationally complex, inefficient, and struggles to guarantee the accuracy of the recognition results. Figures 1a-1c This shows a label image that is difficult to identify using traditional methods.

[0004] The content of the background section is merely the technology known to the inventor and does not necessarily represent the prior art in this field. Summary of the Invention

[0005] To address one or more of the problems existing in the prior art, the present invention provides a label identification method, wherein each label has multiple circular reflective label dots, the identification method comprising:

[0006] Acquire a label image, wherein the label image includes one or more labels;

[0007] Obtain the contour lines of the label image;

[0008] The label points are identified based on the contour lines; and

[0009] Cluster the labeled points and obtain the labels based on the clustering results.

[0010] According to one aspect of the invention, the step of acquiring the label image includes: acquiring the label image using an image acquisition unit mounted on a robot.

[0011] According to one aspect of the present invention, the step of obtaining the contour lines of the label image includes: determining the grayscale values ​​of the pixels of the label image, and obtaining the contour lines based on the grayscale values.

[0012] According to one aspect of the invention, it further includes: determining the number of pixels on each contour line or within the area enclosed by the contour lines; if the number is outside a preset range, excluding the contour lines.

[0013] The step of identifying the label points based on contour lines includes: identifying the label points according to the remaining contour lines.

[0014] According to one aspect of the invention, the step of identifying the label point based on contour lines further includes: determining the roundness of the contour lines, identifying a circular outline based on the roundness, the circular outline corresponding to the outline of the label point.

[0015] According to one aspect of the invention, the step of determining the roundness of the contour line includes: determining the position of the center pixel within the area enclosed by the contour line, and the position of each pixel on the contour line, and determining the distance between each pixel and the center pixel.

[0016] According to one aspect of the invention, the step of determining the roundness of the contour lines further comprises: determining the mean, variance, and standard deviation of the distances, wherein the roundness is related to the mean and standard deviation.

[0017] According to one aspect of the present invention, the step of identifying a circular contour based on roundness includes: when the roundness is greater than a first threshold, determining that the circular contour is the circular contour of the label point.

[0018] According to one aspect of the invention, the step of clustering the label points includes: clustering the label points based on the center position of the label points.

[0019] According to one aspect of the present invention, the step of clustering the label points based on the center position of the label points includes: determining a first center position of a first label point among the label points; finding a second center position of a second label point that is no more than a second threshold away from the first center position; determining the average position of the first center position and the second center position; and searching for the center positions of other label points that are no more than a second threshold away from the average position; and repeating this process until the clustering of all label points is completed.

[0020] The present invention also provides a label identification device, comprising:

[0021] The image acquisition unit is configured to acquire label images; and

[0022] The processing unit is coupled to the image acquisition unit and configured to execute the recognition method as described above.

[0023] The present invention also provides a computer-readable storage medium, including computer-executable instructions and a tag image stored thereon, wherein the executable instructions implement the identification method described above when executed by a processor.

[0024] The present invention also provides a robot, comprising:

[0025] case;

[0026] Mobile chassis with a walking mechanism;

[0027] An image acquisition unit is installed on the robot and configured to acquire tag images;

[0028] The processing unit, coupled to the walking mechanism and the image acquisition unit, is configured to perform the recognition method as described above.

[0029] The technical solution of this invention identifies label points by using contour lines in a label image, thereby identifying the label. Compared with existing technologies, this method is less affected by lighting, produces more accurate identification results, and requires less computation. This improves the efficiency of identification devices and robots using the aforementioned identification method, thereby enhancing the customer experience. Attached Figure Description

[0030] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0031] Figures 1a-1c The image shows a label that is difficult to identify using traditional methods;

[0032] Figure 2 A flowchart illustrating a label identification method according to an embodiment of the present invention is shown;

[0033] Figure 3 A schematic diagram showing a label image including a label according to an embodiment of the present invention is illustrated;

[0034] Figure 4 A schematic diagram showing a label image including multiple labels according to an embodiment of the present invention is illustrated;

[0035] Figure 5a A schematic diagram of contour lines of a label image according to an embodiment of the present invention is shown;

[0036] Figure 5b A schematic diagram of three layers of contour lines according to a preferred embodiment of the present invention is shown;

[0037] Figure 5c It shows Figure 5b A magnified view of the three contour lines shown;

[0038] Figure 6 A schematic diagram illustrating the determination of the roundness of contour lines according to an embodiment of the present invention is shown;

[0039] Figure 7 A schematic diagram of the labels obtained by clustering label points according to a preferred embodiment of the present invention is shown.

[0040] Figure 8 A schematic diagram of a tag identification device according to an embodiment of the present invention is shown; and

[0041] Figure 9 A schematic diagram of a robot according to an embodiment of the present invention is shown. Detailed Implementation

[0042] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0043] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0044] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, an electrical connection, or a connection that allows for communication; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0045] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0046] The following disclosure provides many different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, examples of various specific processes and materials are provided in this invention, but those skilled in the art will recognize the application of other processes and / or the use of other materials.

[0047] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0048] This invention provides a label recognition method based on image contour lines. Compared with existing technologies, it is less affected by lighting conditions and the recognition results are more accurate. The recognition method is described in detail below.

[0049] Figure 2 A flowchart of a tag identification method 100 according to an embodiment of the present invention is shown. Figure 2 As shown, the identification method 100 includes steps S101 to S104. The specific steps of the identification method 100 are described below.

[0050] In step S101, a label image is obtained, the label image including one or more labels.

[0051] The tag image can be acquired by an image acquisition unit installed on a robot, such as a robot in a restaurant, hotel, or library. The image acquisition unit can be, for example, a camera. This invention does not limit the robot's working environment or the specific type of image acquisition unit. Furthermore, this invention does not limit the specific installation location of the image acquisition unit on the robot, or the specific attachment location of the tag in the robot's working environment. The tag can be attached, for example, to the ceiling, floor, or wall in the robot's working environment to guide the robot in positioning or pose correction.

[0052] In some preferred embodiments of the invention, the label can be affixed to the ceiling in the robot's working environment, while the image acquisition unit can be mounted on top of the robot (see reference). Figure 9 This design aims to enable the image acquisition unit to not only acquire the tag image but also reduce the adverse effects of light emitted by other light-emitting devices (such as LiDAR) on the image acquisition unit. It should be understood that the field of view of the image acquisition unit is limited. Therefore, if different tags are far apart, the number of tags in the same tag image will be small; conversely, if different tags are close together, the number of tags in the same tag image will be large. In other words, the tag image may include one or more tags. It should be noted that this invention does not limit the specific number of tags, the spacing between tags, or the size of the tags in the robot's working environment; these can be determined according to the actual situation.

[0053] Figures 3-4 The examples show cases where the label image includes one label and cases where it includes multiple labels. Figure 3 and Figure 4 In the label images shown, each label has multiple circular label dots. In some preferred embodiments, these circular label dots may be made of a reflective material, such as diamond reflective film, etc. The present invention does not limit the specific type of reflective material. In some preferred embodiments, different labels may be affixed to different locations, wherein the number and / or arrangement of the label dots on each label are different. Figure 4 This scenario is illustrated exemplarily. Therefore, in practical applications, robots can identify different tags by recognizing the number and / or arrangement of the tag points, thereby determining the position corresponding to each tag, and thus performing localization or pose correction.

[0054] The above embodiments describe the case where the label dots are made of reflective material. Alternatively, the label dots can also be made of light-emitting devices such as light-emitting diodes (LEDs). The present invention does not limit the choice of material for the label dots; in practical applications, the choice can be made according to the actual situation. It should be understood that the purpose of using either light-emitting or reflective materials for the label dots is to highlight their brightness, so that the label dots can be distinguished from the brightness of other locations on the label or other objects in the label image, thereby enabling efficient and accurate identification of the label dots and ultimately, the label itself (described later). It should be noted that the present invention does not limit the size of the label dots. Preferably, the label dots on the same label are of the same size and material to reduce the complexity and computational load of label recognition, thereby improving the efficiency and accuracy of label recognition.

[0055] In step S102, the contour lines of the label image are obtained.

[0056] According to a preferred embodiment of the present invention, the grayscale values ​​of the pixels in the label image can be determined, and the contour lines can be obtained based on the grayscale values. In practical applications, the label image may include other objects besides the label. Therefore, to highlight the label points, reduce interference from other objects, and reduce computational load, the label image can be grayscale processed. The present invention does not limit the specific method of grayscale processing; for example, any one of the maximum value method, average value method, and weighted average method can be used, depending on the specific situation. After grayscale processing, each pixel in the label image can be traversed to obtain the grayscale value of each pixel. It should be understood that since the label points are made of reflective or luminescent materials, the brightness of the label points will be significantly higher than the brightness of other positions on the label or other objects in the label image. That is, the grayscale values ​​of the pixels occupied by the label points should be significantly higher and approximately equal. Therefore, pixels with approximately equal grayscale values ​​can be connected to form a closed curve, which is the contour line, and the contour line can be used to represent the outline of an object.

[0057] Figure 5a A schematic diagram of the contour lines of a label image according to a preferred embodiment of the present invention is shown. Figure 5b A schematic diagram of three layers of contour lines according to a preferred embodiment of the present invention is shown. Figure 5c It shows Figure 5b A magnified view of the three contour lines shown. (See attached image.) Figures 5a-5c As shown, the circular outline of the label point is clearly visible. Therefore, label recognition can be transformed into the process of recognizing the circular outline. Figures 5b-5cIt can be seen that some contour lines have connected boundaries, especially the outermost layers. This may be due to uneven lighting. Therefore, three layers of contour lines can be selected for processing to avoid connected boundaries and improve the accuracy of subsequent processing. Furthermore, since the label points have a preset size, their outlines (contour lines) should also have a preset size. Therefore, to reduce the complexity and computational load of subsequent processing, contour lines that are clearly not label points can be excluded first. In some preferred embodiments of the present invention, the number of pixels on each contour line or within the area enclosed by the contour lines can be determined. If the number is outside a preset range (i.e., greater than a certain threshold), the contour line is excluded, and the label point is identified based on the remaining contour lines. How to identify label points based on contour lines will be described in detail below.

[0058] In step S103, the label points are identified based on contour lines.

[0059] According to a preferred embodiment of the present invention, the roundness R of a contour line can be determined, and a circular outline can be identified based on the roundness R. The circular outline corresponds to the outline of the label point. It should be noted that the roundness R represents the degree of roundness of the contour line. The larger the value of roundness R, the rounder the contour line. When the roundness R is greater than a first threshold T1, the contour line (circular outline) is determined to be the circular outline of the label point. Therefore, using roundness R as a roundness determination index is simple and intuitive. The method for determining the roundness R of the contour line will be described in detail below.

[0060] Figure 6 A schematic diagram illustrating the determination of contour circleness according to a preferred embodiment of the present invention is shown. Figure 6 As shown, the position P0(x0,y0) of the center pixel within the area enclosed by the contour lines can be determined, and the position P of each pixel on the contour lines can be determined. i (x i ,y i The number of pixels N on the contour lines is determined, and the number of each pixel P is determined. i (x i ,y i The distance Di between the center pixel P0(x0,y0) and the center pixel P0(x0,y0). Preferably, the distance is a Euclidean distance, which can be calculated using the following formula 1-1:

[0061]

[0062] According to a preferred embodiment of the present invention, the step of determining the roundness R of the contour lines further includes: determining the average value Dave and variance S of the distance Di. 2And the standard deviation S, wherein the roundness R is related to the mean Dave and the standard deviation S. The mean Dave can be calculated using the following formulas 1-2:

[0063]

[0064] The variance S 2 The following formulas 1-3 can be used for calculation:

[0065]

[0066] The standard deviation S is the variance S. 2 The arithmetic mean root can be calculated using the following formulas 1-4:

[0067]

[0068] The roundness R can be calculated using the following formulas 1-5:

[0069]

[0070] The above embodiments describe how to determine the roundness R of contour lines. It should be noted that the roundness of contour lines is positively correlated with the value of roundness R; that is, the larger the value of roundness R, the rounder the contour line. According to a preferred embodiment of the present invention, when the roundness R is greater than a first threshold T1, the contour line (circular outline) is determined to be the circular outline of the label point. The specific value of the first threshold T1 can be determined according to the actual situation. In some specific embodiments of the present invention, the first threshold T1 can be 0.8. That is, contour lines with a roundness R greater than 0.8 correspond to the circular outline of the label point, thereby identifying the label point. In some preferred embodiments, the average roundness of the three layers of contour lines can be determined according to the above method. If the average roundness of the three layers of contour lines is greater than the first threshold T1, then these three contour lines correspond to the circular outline of the label point, thereby identifying the label point. After identifying the label point, it is also necessary to determine the label corresponding to the label point, which will be described next.

[0071] In step S104, the label points are clustered, and the labels are obtained based on the clustering results.

[0072] According to a preferred embodiment of the present invention, the label points can be clustered based on their center positions (centers). It should be understood that label points on the same label are closer together, while label points on different labels are farther apart. Therefore, the distances between label points can be determined based on their center positions (centers) to cluster the label points and thus obtain different labels.

[0073] Figure 7A schematic diagram of the labels obtained by clustering label points according to a preferred embodiment of the present invention is shown. Figure 7 As shown, among multiple label points, the first center position Q1(x1,y1) of the first label point can be determined. The second center position Q2(x2,y2) of the second label point is found that is no more than a second threshold T2 away from the first center position Q1(x1,y1). The average position Q(x,y) of the first center position Q1(x1,y1) and the second center position Q2(x2,y2) is determined. The center positions Qi(xi,yi) of other label points that are no more than a second threshold T2 away from the average position Q12 are then searched. This process is repeated until all label points are clustered, resulting in multiple labels (see reference). Figure 7 (The portion enclosed by a solid rectangular line). It should be noted that the center position can be the center pixel, and the first label point can be any one of multiple label points; this invention does not impose any limitations. The average position can be calculated using the following formula (1-6):

[0074]

[0075] It should be noted that the present invention does not impose any limitations on the specific value of the second threshold T2. Furthermore, the first threshold T1 and the second threshold T2 are not necessarily related in magnitude, but depend on the specific circumstances.

[0076] The label recognition method 100 has been described in detail above. This method identifies the circular outline of label points based on the roundness of contour lines, thereby recognizing the label. It is less affected by lighting conditions and also less affected by the label points themselves. Regardless of the size of the label points, it can accurately identify them and significantly reduces computational complexity and workload, thus improving recognition efficiency. Using the recognition method 100 of this invention, the following tasks can be easily handled: Figures 1a-1c This illustrates a case where a label image is difficult to identify using traditional methods.

[0077] It should be noted that identifying the circular outline of the label points based on contour lines is a preferred embodiment of the present invention. In practical applications, in addition to identifying the circular outline based on contour lines, traditional image processing methods can also be used, including but not limited to Hough transform, smoothing filtering, threshold calculation, binarization, etc. By filtering out scattered noise points, calculating the threshold of the grayscale histogram, performing binarization processing based on the threshold, and applying nearest-point clustering to form labels from light spots, all of these are within the protection scope of the present invention.

[0078] The present invention also provides a label identification device 200. Figure 8 A schematic diagram of a tag identification device 200 according to an embodiment of the present invention is shown, such as... Figure 8 As shown, the identification device 200 includes an image acquisition unit 30 and a processing unit 40, wherein the image acquisition unit 30 is configured to acquire a label image; the processing unit 40 is coupled to the image acquisition unit 30 and is configured to execute the identification method 100 as described above.

[0079] The present invention also provides a robot 300. Figure 9 A schematic diagram of a robot 300 according to an embodiment of the present invention is shown. Figure 9 As shown, the robot 300 includes: a housing 20, a mobile chassis 10, an image acquisition unit 30, and a processing unit (not shown in the figure). The mobile chassis 10 has a walking mechanism; the image acquisition unit 30 is mounted on the robot 300 and configured to acquire label images; the processing unit (not shown in the figure) is coupled to the walking mechanism and the image acquisition unit 30 and configured to execute the recognition method 100 described above.

[0080] According to a preferred embodiment of the present invention, the image acquisition unit 30 may be installed on the top of the robot 300 to acquire tags located on the ceiling. The image acquisition unit 30 may be, for example, a camera.

[0081] According to a preferred embodiment of the present invention, a plurality of sensors are also installed on the robot 300 to detect the surrounding environment of the robot 300. The plurality of sensors include one or more of the following: a lidar 50, a binocular vision camera, an odometer, a stereo vision sensor, and an infrared sensor.

[0082] It should be understood that in practical applications, robots may be equipped with various sensors (such as lidar, odometers, etc.) for positioning. However, as the robot's working time increases, or due to interference from the surrounding environment, the sensors may accumulate errors, leading to discrepancies between the sensor data and the actual data. This results in the robot being unable to accurately position itself, thus affecting its normal operation. In some preferred embodiments of the present invention, the tag may carry location information (such as the number and / or arrangement of tag points) and may be fixedly attached to a fixed position in the robot's working environment. Therefore, the robot can periodically acquire tag images through the image acquisition unit and identify the location information carried by the tag for positioning and error compensation, thereby achieving pose correction and improving positioning accuracy.

[0083] According to a preferred embodiment of the present invention, the lidar 50 can be disposed at the opening of the housing 20, thereby making it easy for the lidar 50 to emit laser signals to detect the surrounding environment. The lidar 50 can be a multi-line lidar or a single-line lidar, which can be selected according to specific circumstances in practical applications. When the lidar 50 is a multi-line lidar, the lidar 50 includes a photoelectric receiving array and a laser emitting unit array, so that when the lidar 50 rotates along the set plane, the photoelectric receiving array can form a scanning cylinder, thereby increasing the scanning area, facilitating the identification of obstacle shape details, and reducing the occurrence of collisions. When the lidar 50 is a single-line lidar, the lidar 50 only contains a single photoelectric receiving unit and a single laser emitting unit. After the lidar 50 rotates along the set plane, it can only measure the shape of an object in a circle, and cannot timely acquire the shape of complex objects, which is prone to collisions and endangers personal and property safety. Therefore, it can be used in conjunction with other sensors such as binocular vision cameras, odometers, stereo vision sensors, and infrared sensors to detect the surrounding environment. It should be noted that the aforementioned set plane can be a horizontal plane, which facilitates the robot's object detection during movement. Of course, other set planes, such as a vertical plane, can also be selected as needed, and this invention does not limit this.

[0084] According to a preferred embodiment of the present invention, the bottom of the mobile chassis 10 is provided with at least one turn signal unit 110, and each turn signal unit 110 includes at least one turn signal 111; the walking mechanism is provided with at least two sets of drive wheels 120, each set of drive wheels 120 being located on one side of the mobile chassis 10; the component controller controls the traveling speed of the drive wheels 120; and controls the turn signals 111 in the turn signal unit 110 to illuminate in a preset manner when the robot turns, so as to alert pedestrians. Among the drive wheels 120, at least one set of drive wheels 120 is used as the left drive wheel, and at least one set of drive wheels 120 is used as the right drive wheel, with the left and right drive wheels located on opposite sides of the mobile chassis 10. Optionally, the walking mechanism may also include at least two sets of driven wheels, one set of driving wheels corresponding to one set of driven wheels, wherein at least one set of driven wheels is used as the left driven wheel, and at least one set of driven wheels is used as the right driven wheel. The left and right driven wheels are used to assist the left and right driving wheels in driving the robot's housing 20 and mobile chassis 10 to move, so as to reduce the load pressure on the driving wheel 120.

[0085] Based on the above technical solution, optionally, when the speed difference between the drive wheels 120 on both sides of the mobile chassis 10 is greater than a preset value, the component controller controls the turn signal 111 in the turn signal unit 110 to light up in a preset manner.

[0086] Optionally, the robot 300 may also include a voice module (not shown in the figure), which is electrically connected to the component controller. When the robot turns, the component controller can control the voice module to issue voice prompts to announce the tag recognition results and the corresponding position of the tag, thereby assisting the robot in positioning or pose correction. In addition, the voice module can also alert pedestrians or other robots to ensure the safety of pedestrians and robots.

[0087] The present invention also provides a computer-readable storage medium, including computer-executable instructions and a tag image stored thereon, wherein the executable instructions, when executed by a processor, implement the identification method 100 described above. The computer-readable storage medium may be, for example, a portable computer disk, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory stick, floppy disk, etc. The present invention does not limit the specific type of computer-readable storage medium; one or more suitable types can be selected according to actual circumstances.

[0088] The technical solution of this invention identifies label points by using contour lines in a label image, thereby identifying the label. Compared with existing technologies, this method is less affected by lighting, produces more accurate identification results, and requires less computation. This improves the efficiency of identification devices and robots using the aforementioned identification method, thereby enhancing the customer experience.

[0089] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying labels, wherein each label has multiple circular reflective label dots, the identification method comprising: Acquire a label image, wherein the label image includes one or more labels; Obtaining the contour lines of the label image includes: determining the grayscale values ​​of the pixels in the label image, and obtaining the contour lines based on the magnitude of the grayscale values; the contour lines are used to represent the outline of an object. The label points are identified based on the contour lines; and Cluster the labeled points and obtain the labels based on the clustering results.

2. The identification method according to claim 1, wherein the step of acquiring the label image includes: The label image is acquired using an image acquisition unit mounted on the robot.

3. The identification method of claim 1, further comprising: Determine the number of pixels on each contour line or within the area enclosed by the contour lines. If the number is outside a preset range, exclude the contour lines. The step of identifying the label points based on the contour lines includes: identifying the label points based on the remaining contour lines.

4. The identification method of claim 1, wherein the step of identifying the tag point based on the contour further comprises: The roundness of the contour lines is determined, and a circular outline is identified based on the roundness, the circular outline corresponding to the outline of the label point.

5. The method of claim 4, wherein the step of determining the roundness of the contour comprises: Determine the position of the center pixel within the area enclosed by the contour lines, as well as the position of each pixel on the contour lines, and determine the distance between each pixel and the center pixel.

6. The identification method of claim 5, wherein the step of determining the roundness of the contour further comprises: The mean, variance, and standard deviation of the distance are determined, and the roundness is related to the mean and standard deviation.

7. The identification method of claim 6, wherein the step of identifying a circular profile based on the roundness comprises: When the roundness is greater than the first threshold, the circular outline is determined to be the circular outline of the label point.

8. The identification method of any one of claims 1-7, wherein the step of clustering the tag points comprises: The label points are clustered based on their center positions.

9. The method of claim 8, wherein the step of clustering the tag points based on the center locations of the tag points comprises: Among the labeled points, a first center position of the first labeled point is determined, a second center position of the second labeled point that is no more than a second threshold away from the first center position is found, the average position of the first center position and the second center position is determined, and the center positions of other labeled points that are no more than a second threshold away from the average position are searched. This process is repeated until the clustering of all labeled points is completed.

10. The identification method according to any one of claims 1-7, wherein the tag point may be made of a luminescent material or a reflective material.

11. A tag identification device, comprising: The image acquisition unit is configured to acquire label images; and The processing unit is coupled to the image acquisition unit and configured to perform the recognition method as described in any one of claims 1-10.

12. A computer-readable storage medium comprising computer-executable instructions stored thereon and a tag image, wherein the executable instructions, when executed by a processor, implement the identification method as described in any one of claims 1-10.

13. A robot comprising: case; Mobile chassis with a walking mechanism; An image acquisition unit is installed on the robot and configured to acquire tag images; The processing unit, coupled to the walking mechanism and the image acquisition unit, is configured to perform the recognition method as described in any one of claims 1-10.

Images