Image processing method and device, electronic equipment and storage medium
By segmenting areas for the target picture and assigning addressing identifiers, the problem that robots find it difficult to accurately locate target objects in complex environments is solved, and the target objects are quickly and accurately found, which improves work efficiency.
Patent Information
- Application Number
- CN202411017346.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-26
AI Technical Summary
When performing tasks, existing robots are unable to accurately locate target objects, resulting in the need to search for larger areas in complex environments, which reduces work efficiency.
By dividing the target screen into multiple cutting areas and assigning a unique addressing identifier to each area, after identifying the target object, the addressing identifier of the area where the target object is located is determined based on the intersection data of the target detection box and the cutting area, and guide The robot quickly and accurately finds the target object.
Avoid blind search, significantly save time, and improve the work efficiency and accuracy of the robot.
Smart Images

Figure CN119027487A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence, and in particular to an image processing method, device, electronic equipment and storage medium. Background Art
[0002] In the current development of robot technology, especially in the application fields of service robots such as sanitation robots and food delivery robots, achieving accurate positioning of target objects is the key to improving the work efficiency of robots. However, in existing solutions, when the target object (such as garbage) is identified, only a rough indication of the entire image area can be obtained, and the precise location of the target object cannot be directly provided. This results in the robot needing to search and process a larger area when performing tasks, especially in complex environments, such as when multiple target objects are close to or overlapped, which reduces the robot's work efficiency. Summary of the invention
[0003] The embodiment of the present invention provides an image processing method, which aims to solve the problem that the existing robot needs to search and process a large area when performing tasks, which reduces the work efficiency of the robot. By assigning a unique addressing identifier to each cutting area and locating the target object to the specific cutting area after identifying it, the robot can directly find the target object quickly and accurately according to the target addressing identifier, avoiding blind searching in the entire target image, thereby greatly saving time and improving the work efficiency of the robot.
[0004] In a first aspect, an embodiment of the present invention provides an image processing method, the method comprising the following steps:
[0005] Acquire multiple cut areas of the target image, each cut area corresponds to an addressing identifier;
[0006] When a target object is identified in the target picture, a target detection frame of the target object is obtained, wherein the target object is an object that needs to be processed by the robot;
[0007] Determining the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas;
[0008] The addressing mark of the area where the target object is located is determined as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
[0009] Optionally, the acquiring of multiple cut regions of the target image includes:
[0010] Determining a root area in a current cutting process in the target picture;
[0011] In the current cutting process, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current cutting regions of the target image;
[0012] If the cutting is not completed, determining the root area for the next cutting process based on the multiple current cutting areas;
[0013] If the cutting is completed, determining a plurality of cutting areas of the target image based on the plurality of current cutting areas;
[0014] Each cutting area is assigned an addressing identifier.
[0015] Optionally, in the current sub-cutting process, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current sub-cutting regions of the target image, including:
[0016] In the current cutting process, obtaining the cutting straight line;
[0017] Performing a region cutting process on the root region by using the cutting straight line, and determining whether the number of intersections between the cutting straight line and the root region is two;
[0018] If the number of intersections between the cutting line and the root region is not two, the region cutting process is determined to be invalid, and the cutting line is abandoned;
[0019] If there are two intersection points between the cutting line and the root area, the area cutting process is determined to be valid, and the cutting line is retained to obtain a plurality of current sub-cutting areas of the target image.
[0020] Optionally, the retaining of the cutting straight line to obtain a plurality of current sub-cutting areas of the target image includes:
[0021] Obtain a first vertex set of the root region, wherein the first vertex set includes a plurality of first vertices including a first starting vertex and a first ending vertex, the first vertices being endpoints of edges of the root region, and the plurality of first vertices being arranged in a first set order, which is a clockwise order or a counterclockwise order of the root region;
[0022] Obtain two intersection points of the cutting line and the root region, and insert the two intersection points into the vertex set in the order of the first set to obtain a second vertex set;
[0023] Based on the second vertex set, a plurality of current sub-cutting areas of the target picture are determined.
[0024] Optionally, the determining, based on the second vertex set, a plurality of current sub-cutting areas of the target image includes:
[0025] In the second vertex set, a plurality of arrays are determined in the order of the first set;
[0026] Based on the multiple arrays, multiple current sub-cutting regions of the target picture are determined, and each array corresponds to one current sub-cutting region.
[0027] Optionally, assigning an addressing identifier to each cutting area includes:
[0028] Determining the centroid coordinates of each of the cutting areas;
[0029] Based on the centroid coordinates of each cutting area, an addressing identifier is assigned to each cutting area.
[0030] Optionally, determining the centroid coordinates of each cutting area includes:
[0031] Obtain a set of vertices to be calculated corresponding to each cutting area, wherein the set of vertices to be calculated includes a plurality of second vertices including a second starting vertex and a second ending vertex of the cutting area, wherein the second vertices are endpoints of edges of the cutting area, and the plurality of second vertices are arranged in a second set order, wherein the second set order is a clockwise order or a counterclockwise order of the cutting area, and each cutting area corresponds to one set of vertices to be calculated;
[0032] Based on the set of vertices to be calculated, determine the directed area between the current second vertex and the remaining second vertices, where the directed area is the vector cross product area between the current second vertex and the next second vertex;
[0033] Based on the directed area, determining the weighted accumulated coordinate value of the cutting area and the accumulated directed area of the cutting area;
[0034] Based on the weighted accumulated coordinate values and the accumulated directed area, the centroid coordinates of the cutting area are determined.
[0035] Optionally, determining the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas includes:
[0036] In the target picture, extending the reference edge of the target detection frame toward one end to obtain a reference straight line of the target detection frame;
[0037] The number of intersections between the reference straight line and each of the cutting areas is calculated, and the area where the target object is located is determined in the cutting area.
[0038] In a second aspect, an embodiment of the present invention further provides an image processing device, the image processing device comprising:
[0039] A first acquisition module is used to acquire multiple cut areas of the target image, each cut area corresponds to an addressing identifier;
[0040] A second acquisition module is used to acquire a target detection frame of the target object when a target object is identified in the target picture, wherein the target object is an object that needs to be processed by the robot;
[0041] A first processing module, configured to determine the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas;
[0042] The second processing module is used to determine the addressing mark of the area where the target object is located as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
[0043] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the image processing method provided in the embodiment of the present invention when executing the computer program.
[0044] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the image processing method provided in the embodiment of the invention are implemented.
[0045] In an embodiment of the present invention, multiple cutting areas of a target screen are obtained, and each cutting area corresponds to an addressing identifier; when a target object is identified in the target screen, a target detection frame of the target object is obtained, and the target object is an object that needs to be handled by the robot; according to the intersection data of the target detection frame and each cutting area, the area where the target object is located is determined in the cutting area; the addressing identifier of the area where the target object is located is determined as the target addressing identifier, and the target addressing identifier is used to guide the robot to find the area where the target object is located. By assigning a unique addressing identifier to each cutting area, and after identifying the target object, locating it to a specific cutting area, the robot can directly find the target object quickly and accurately according to the target addressing identifier, avoiding blind search in the entire target screen, thereby greatly saving time and improving the work efficiency of the robot. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0047] Figure 1 is a flow chart of an image processing method provided by an embodiment of the present invention;
[0048] Figure 2 is a schematic diagram of an uncut target image provided by an embodiment of the present invention;
[0049] Figure 3 is a schematic diagram of a target image after the first cutting provided by an embodiment of the present invention;
[0050] Figure 4 is a schematic diagram of a target image after cutting is completed according to an embodiment of the present invention;
[0051] Figure 5 is a structural schematic diagram of an image processing device provided in an embodiment of the present invention;
[0052] Figure 6 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.
[0054] like Figure 1 As shown, Figure 1 : is a flowchart of an image processing method provided by an embodiment of the present invention, the image processing method comprising the steps of:
[0055] 101. Acquire multiple cutout regions of a target image.
[0056] In an embodiment of the present invention, the image processing method can be used in an urban governance platform. The urban governance platform can be built based on a server or a distributed server. The urban governance platform is deployed with monitoring device interfaces, display screens, target detection models or target detection model interfaces and other functional applications that support image processing. The urban governance platform communicates with a number of monitoring devices through the monitoring device interface to obtain real-time monitoring screens of each monitoring device.
[0057] The above-mentioned urban governance platform is also deployed with a robot interface, through which communication with the robot is performed. The above-mentioned robot may be a sanitation robot, a food delivery robot or other types of robots. The above-mentioned robot is used to handle events, and the above-mentioned events may be garbage cleaning events, food delivery events, etc. Different types of robots handle different types of events. The embodiments of the present invention are mainly described with sanitation robots. For food delivery robots or other types of robots, the description of sanitation robots can be referred to under the purpose of the embodiments of the present invention.
[0058] The target screen may be a monitoring screen of a monitoring device specified by a user, or a monitoring screen of a default monitoring device. It should be noted that each monitoring device corresponds to a monitoring screen, and the monitoring screen specified by a user or defaulted is the target screen.
[0059] A cutting area is correspondingly set in the target image. The cutting area can be a small area divided from the target image, and each cutting area corresponds to an addressing identifier. The cutting area can be obtained by automatically cutting the target image or manually cutting the target image. The addressing identifier is a unique identifier of the cutting area. Each addressing identifier can correspond to a real coordinate or a space coordinate of the robot, so that the robot can convert the addressing identifier into a real coordinate or a space coordinate of the robot.
[0060] In a possible embodiment, the area of the cutting region in the real world may be greater than or equal to the body area of the robot in the real world.
[0061] 102. When a target object is identified in the target image, a target detection frame of the target object is obtained.
[0062] In the implementation of the present invention, the target detection model can be used to perform target detection on the target picture, and then detect whether there is a target object in the target picture. The above-mentioned target object is an object that needs to be disposed of by a robot. For example, the object that needs to be disposed of by a sanitation robot is garbage, that is, the target object at this time is garbage. The target detection model performs target detection on the target picture, and then detects whether there is garbage in the target picture.
[0063] If the target object is not detected, the detection will continue for the following images. If the target object is detected, the target detection model will automatically output a target detection frame, which is the area of the target object in the target image and can be represented by (x, y, w, h), where (x, y) is the representative point of the target detection frame, which can be the center point of the target detection frame or the vertex of the target detection frame, w is the width of the target detection frame in the target image, and h is the height of the target detection frame in the target image.
[0064] 103. According to the intersection data between the target detection frame and each cutting area, the area where the target object is located is determined in the cutting area.
[0065] In the embodiment of the present invention, the intersection data between the target detection frame and each cutting area can be calculated to determine in which cutting area the target object is located, and the cutting area is determined as the area where the target object is located.
[0066] The number of intersections between the extended line of the target detection frame and each cutting area can be calculated. If the number of intersections is an even number, it is determined that the target detection frame is not in the corresponding cutting area, and the cutting area can be determined not to be the area where the target object is located. If the number of intersections is an odd number, it is determined that the target detection frame is in the corresponding cutting area, and the cutting area can be determined to be the area where the target object is located.
[0067] In a possible embodiment, it is possible to calculate in which cutting area the coordinates of the center point of the target detection frame fall, and determine the cutting area where the coordinates of the center point fall as the area where the target object is located.
[0068] In a possible embodiment, the intersection area between the target detection frame and the surrounding cutting areas may be calculated, and the cutting area with the largest intersection area may be determined as the area where the target object is located.
[0069] 104. Determine the addressing identifier of the area where the target object is located as the target addressing identifier, and the target addressing identifier is used to guide the robot to find the area where the target object is located.
[0070] In the embodiment of the present invention, each cutting area corresponds to an addressing identifier. Based on the corresponding relationship between the cutting area and the addressing identifier, after the area where the target object is located is determined, the addressing identifier of the area can be determined as the target addressing identifier.
[0071] Each addressing identifier corresponds to a real coordinate or the spatial coordinate of the robot. After the target addressing identifier is determined, the corresponding real coordinate or the spatial coordinate of the robot can be matched for the target addressing identifier according to the correspondence between the addressing identifier and the real coordinate or the spatial coordinate of the robot, and the matched real coordinate or the spatial coordinate of the robot is sent to the robot. The robot converts the received real coordinate or the spatial coordinate of the robot into a task route, reaches the area where the target object is located according to the task route, and disposes of the target object in the area where the target object is located.
[0072] In an embodiment of the present invention, multiple cutting areas of a target screen are obtained, and each cutting area corresponds to an addressing identifier; when a target object is identified in the target screen, a target detection frame of the target object is obtained, and the target object is an object that needs to be handled by the robot; according to the intersection data of the target detection frame and each cutting area, the area where the target object is located is determined in the cutting area; the addressing identifier of the area where the target object is located is determined as the target addressing identifier, and the target addressing identifier is used to guide the robot to find the area where the target object is located. By assigning a unique addressing identifier to each cutting area, and after identifying the target object, locating it to a specific cutting area, the robot can directly find the target object quickly and accurately according to the target addressing identifier, avoiding blind search in the entire target screen, thereby greatly saving time and improving the work efficiency of the robot.
[0073] It can be understood that in the specific implementation of this application, it involves screen data, image data and other related data. When the embodiments in this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data, as well as the training, deployment and calling of the model, must comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0074] Optionally, in the step of acquiring multiple cutting areas of the target picture, the root area in the current cutting process can be determined in the target picture; in the current cutting process, the root area is subjected to regional cutting processing by using the acquired cutting straight line to obtain multiple current cutting areas of the target picture; if the cutting is not completed, the root area for the next cutting process is determined based on the multiple current cutting areas; if the cutting is completed, the multiple cutting areas of the target picture are determined based on the multiple current cutting areas; and an addressing identifier is assigned to each cutting area.
[0075] In an embodiment of the present invention, when obtaining multiple cutting areas of the target screen, it can be achieved by cutting and marking step by step. First, an initial root area can be determined. The initial root area can be an area drawn by the user or an area obtained according to the image segmentation technology. The root area is the area where the robot performs the disposal task. Then, the area is cut by the received cutting line to obtain a series of sub-areas. The cutting line can be a straight line drawn by the user or automatically generated. Finally, the cutting process will be cyclically performed until the conditions for cutting are met, and finally multiple cutting areas of the target screen are obtained, and a unique addressing identifier is assigned to each area.
[0076] Specifically, the above-mentioned cutting process can be performed multiple times, and each time the root area is cut by a cutting line, and the root area will be reselected in each cutting process. After each cutting process is completed, a unique identifier of the corresponding cutting area can be generated, and the unique identifier is used to distinguish different cutting areas. The above-mentioned unique identifier can be a number or an addressing identifier.
[0077] More specifically, if the cutting is not completed, one or more sub-regions are selected as the root region for the next cutting process based on the multiple sub-regions obtained by the current cutting. The cutting process will be repeated until the cutting completion condition is met. The above-mentioned cutting completion condition can be the number of cutting times or the cutting accuracy. The cutting accuracy can be expressed by the ratio between the area of the cutting area in the real world and the area of the robot. For example, for a root region, the cutting accuracy can be N times the area of the cutting area in the real world and the area of the robot. N is a value between 1 and 2, and the cutting process of the root region is completed.
[0078] When selecting multiple sub-regions as the root region for the next cutting process, the multiple sub-regions can be merged to obtain the merged region as the root region for the next cutting process. In this way, the root region can have a clear boundary. At the same time, it can also facilitate the next cutting process, so that the next cutting process can be determined based on the two intersection points.
[0079] In some possible embodiments, the conditions for completing the cutting may also be set according to actual needs, such as reaching a certain cutting depth, forming enough sub-regions, or satisfying specific image features.
[0080] By gradually cutting and marking the target image, the target image area can be divided more finely and each area can be assigned a unique addressing identifier, thereby improving the accuracy and efficiency of image processing and providing the robot with more accurate location information, thereby significantly improving the robot's work efficiency and accuracy.
[0081] Optionally, in the step of performing regional cutting processing on the root area through the obtained cutting line to obtain multiple current sub-cutting areas of the target picture, in the current sub-cutting processing, the cutting line can be obtained; the root area is subjected to regional cutting processing through the cutting line, and it is determined whether the number of intersection points between the cutting line and the root area is two; if the number of intersection points between the cutting line and the root area is not two, the regional cutting processing is determined to be invalid, and the cutting line is abandoned; if the number of intersection points between the cutting line and the root area is two, the regional cutting processing is determined to be valid, the cutting line is retained, and multiple current sub-cutting areas of the target picture are obtained.
[0082] In the embodiment of the present invention, the cutting straight line can be a straight line drawn by a user or automatically generated (usually understood as a line segment with two endpoints). Specifically, a straight line for cutting can be selected or generated according to a certain algorithm or input by the user on the target screen, and the starting point and end point of the straight line are determined.
[0083] Use the above cutting line to cut the current root area. During the cutting process, the intersection of the cutting line and the root area can be calculated. A root area can also be understood as a polygon. The intersection of the cutting line and the root area is the intersection of the cutting line and the upper edge of the corresponding polygon.
[0084] If there are not two intersection points between the cutting line and the root area, it means that the line is outside the root area, or the line is inside the root area (in this case, the line is shorter), then it can be determined that the current cutting process is invalid. If the cutting process is invalid, the current cutting line can be abandoned. A new cutting line can be regenerated later to continue the cutting process.
[0085] If there are two intersection points between the cutting line and the root area, it can be judged that the cutting is valid. At this time, the cutting line can be retained and the root area can be divided into two new sub-areas according to the positions of the intersection points.
[0086] After confirming that the cutting process is effective, two new polygon point sets can be generated according to the intersection of the cutting line and the root area, representing two new sub-areas respectively. These two sub-areas can be assigned their own identifiers as the cutting areas of the root area. It can be understood that since each cutting is performed only through one cutting line, after the cutting process of the root area is performed, the cutting area of the target image in the current cutting process can also be obtained.
[0087] The root region is effectively cut by cutting straight lines, thereby generating multiple smaller sub-regions as cutting regions, which can improve the accuracy of the cutting region in image processing.
[0088] Specifically, the following method can be used to check whether they are two intersection points: given the endpoint coordinates of each edge in the root area and the starting and ending points of the cutting line, each edge of the root area is traversed to calculate whether it intersects with the cutting line. The specific calculation method is as follows: given the two points of the polygon edge (x1, y1, x2, y2) and the two points of the cutting line (x3, y3, x4, y4), the denominator of the two lines is calculated. The formula is: denominator = (y4-y3) × (x2-x1) - (x4-x3) × (y2-y1). If the denominator is 0, it means that the two lines are parallel or coincident. Calculate the parameters ua and ub. ua and ub are used to determine the position of the intersection on each line segment. The formula is as follows: ua = ((x4-x3) × (y1-y3) - (y4-y3) × (x1-x3)) / denominator; ub = ((x2-x1) × (y1-y3) - (y2-y1) × (x1x3)) / denominato. If ua and ub are both between 0 and 1 (including 0 and 1), it means that the intersection is on two line segments, and calculate the coordinates of the intersection intersectionX and intersectionY. The formula is as follows: intersectionX = x1 + ua × (x2-x1), intersectionY = y1 + ua × (y2-y1). After the traversal is completed, determine the number of intersections. If there are 2, it is a valid cut.
[0089] In a possible embodiment, the cutting lines corresponding to the effective cutting processes can be retained as effective cutting lines, and the cutting line set includes all effective cutting lines when the cutting is completed. After all the cutting processes are completed, all effective cutting lines can be drawn in the target screen to obtain multiple cutting areas.
[0090] Optionally, in the step of retaining the cutting line to obtain multiple current sub-cutting areas of the target image, a first vertex set of the root area can be obtained, the first vertex set includes multiple first vertices including a first starting vertex and a first ending vertex, the first vertex is the endpoint of the edge of the root area, and the multiple first vertices are arranged in a first set order, and the first set order is a clockwise or counterclockwise order of the root area; obtain two intersection points of the cutting line and the root area, insert the two intersection points into the vertex set in the first set order to obtain a second vertex set; based on the second vertex set, determine multiple current sub-cutting areas of the target image.
[0091] In an embodiment of the present invention, the first vertex set of the root area includes the set of endpoints of all edges of the root area. The root area is a polygon. Therefore, the endpoints of the edges of the root area can also be called vertices of the polygon. The above-mentioned first vertices are arranged in a set order (clockwise or counterclockwise) so that the shape of the polygon (cutting area) can be accurately constructed in subsequent processing.
[0092] When performing the cutting process, the intersection point of the cutting line and the root area can be calculated. The intersection point represents the intersection position of the cutting line and the boundary of the root area.
[0093] It should be noted that there must be two intersections between the cutting line and the root area to be considered as a valid cutting process. After being determined as a valid cutting process, the two intersections can be inserted into the set in the order of the first set of the first vertex set of the root area, so that the combination of the cutting area can be ensured to be consistent with the original root area in shape and structure, and also include the new vertices and new edges introduced by the cutting line.
[0094] After the intersection is inserted into the first vertex set, the first vertex set is updated to the second vertex set. The second vertex set contains all the vertices required to construct the cutting area, including all the first vertices of the root area, the two intersections of the cutting lines, and other necessary first vertices. Other necessary first vertices may be repeated first vertices added to close the polygon.
[0095] After obtaining the second vertex set, multiple new polygons can be constructed, namely the current cutting area. The current cutting area is formed by connecting the first vertex and two intersection points in the second vertex set, and can represent the parts of the root area after being divided by the cutting line.
[0096] The cutting accuracy of the cutting area can be improved by obtaining the first vertex set of the root area, calculating the intersection of the cutting line and the root area, inserting the intersection into the first vertex set to obtain the second vertex set, and finally determining multiple current sub-cutting areas of the target image based on the second vertex set.
[0097] Optionally, in the step of determining multiple current sub-cutting areas of the target image based on the second vertex set, multiple arrays can be initialized; in the second vertex set, multiple arrays are determined in the order of the first set; based on the multiple arrays, multiple current sub-cutting areas of the target image are determined, each array corresponds to one of the current sub-cutting areas.
[0098] In an embodiment of the present invention, a first array can be initialized; in a second vertex set, the first starting vertex and all first vertices between the first starting vertex and the first intersection are added to the first array in sequence according to the order of the first set, the first intersection and the second intersection are added to the first array in sequence, and all first vertices from the second intersection to the first ending vertex are added to the first array in sequence, and the first starting vertex is the same as the first ending vertex; a second array is initialized; in the second vertex set, the first intersection is added to the second array in sequence according to the order of the first set, all first vertices between the first intersection and the second intersection are added to the second array in sequence, and the second intersection is added to the second array; based on the first array and the second array, multiple current sub-cutting areas of the target screen are determined, and each array corresponds to a current sub-cutting area.
[0099] In the embodiment of the present invention, for the current cutting process, a data structure of a first array can be initialized, the first array is used to store the vertices required to construct the first current cutting area, and a data structure of a second array is initialized, the second array is used to store the vertices required to construct the second current cutting area.
[0100] In the second vertex set, you can traverse in the order of the first set. First, add all the first vertices from the first starting vertex to the first intersection to the first array in sequence, which ensures that part of the boundary of the cutting area is consistent with the part of the boundary of the corresponding root area. Next, add the first intersection and the second intersection to the first array in sequence. These two intersections are the endpoints of the new edges of the cutting area. Finally, add all the first vertices from the second intersection to the first ending vertex to the first array in sequence to complete the boundary closure of the cutting area.
[0101] Similar to the filling process of the first array, in the second vertex set, first add the first intersection point to the second array as the starting point of the cutting area. Then, add all the first vertices between the first intersection point and the second intersection point to the second array in sequence. These first vertices represent the common part of the boundary between the cutting area and the root area. Finally, add the second intersection point to the end of the second array.
[0102] Specifically, after the traversal calculation is completed, the number of intersections is determined. If the number of intersections is 2, it is a valid cut. At the same time, in the process of traversing and calculating the intersections, record the intersection formed by the intersection of the cutting line and the edge of the root area, and sort the intersection as index+1. After completing the above steps, the coordinates of the two intersections and the sorting of each intersection are known, and the polygon endpoint set (the second vertex set of the root area) can be constructed. The process of constructing the cutting area is as follows: Initialize an empty array s1Points (the first array), add all vertices from the starting point to the first intersection in the original polygon to s1Points, add the coordinates of the first intersection and the second intersection to s1Points, and add all vertices from the second intersection to the end point in the original polygon to s1Points. Initialize an empty array s2Points (the second array), add the coordinates of the first intersection to s2Points, add all vertices from the first intersection to the second intersection in the original polygon to s2Points, and add the coordinates of the second intersection to s2Points. In this way, two new polygon point sets are formed, each of which represents a vertex coordinate set of a cutting area.
[0103] By initializing the array, filling the first array and the second array in a specific order, and finally determining multiple current sub-cutting areas of the target picture based on the first array and the second array, the cutting accuracy and cutting speed of the current sub-cutting area can be improved.
[0104] In a possible embodiment, the cutting straight line includes a first type of cutting straight line and a second type of cutting straight line, wherein any two first type of cutting straight lines do not intersect in the target screen, any two second type of cutting straight lines do not intersect in the target screen, there is at most one intersection between a first type of cutting straight line and a second type of cutting straight line, and the intersection between the first type of cutting straight line and the second type of cutting straight line is the middle intersection. The first type of cutting straight line has a first direction, and the second type of cutting straight line has a second direction. For example, the first direction can be a horizontal direction, and the second direction is a vertical direction, or the first direction can be a vertical direction, and the second direction is a horizontal direction. It should be noted that the horizontal direction does not refer to an absolute horizontal direction, but a horizontal direction with a certain error angle, and the vertical direction does not refer to an absolute vertical direction, but a vertical direction with a certain error angle. The root area can be obtained by merging multiple sub-areas in the corresponding direction of the cutting straight line in the current cutting process. For example, if the cutting straight line is in the first direction, it is obtained by merging multiple sub-areas in the first direction, and there is one sub-area in the second direction. The root area can be understood as a row or a column.
[0105] For the current cutting process, a data structure of a first array can be initialized, and the first array is used to store the vertices required to construct the first current cutting area. A data structure of a second array is initialized, and the second array is used to store the vertices required to construct the second current cutting area. And so on, there can be n arrays.
[0106] The second vertex set may include a plurality of first vertices including a first starting vertex and a first ending vertex, as well as a first historical intersection point and a second historical intersection point of a historical effective cutting line and a root area before the current cutting process. The traversal may be performed in the order of the first set. First, all first vertices between the first starting vertex and the first intersection point are sequentially added to the first array, thereby ensuring that a portion of the boundary of the cutting area is consistent with a portion of the boundary of the corresponding root area.
[0107] Next, calculate whether the new edge connected by the first intersection point and the second intersection point has an intersection with the valid cutting line. If the new edge connected by the first intersection point and the second intersection point does not have an intersection with the historical valid cutting line, add the first intersection point and the second intersection point to the first array in sequence. These two intersection points are the endpoints of the new edge of the cutting area. In the second vertex set, first, add the first intersection point to the second array as the starting point of the cutting area. Then, add all the first vertices between the first intersection point and the second intersection point to the second array in sequence. These first vertices represent the common part of the boundary between the cutting area and the root area. Finally, add the second intersection point to the end of the second array.
[0108] If the new edge connected by the first intersection and the second intersection has an intersection with the historical effective cutting line, the intersection is determined as the middle intersection. The first intersection, the middle intersection and the first historical intersection are added to the first array in the order of the first set; the first intersection, all the first vertices between the first intersection and the second historical intersection, the second historical intersection, and the middle intersection are added to the second array in the order of the first set; the second historical intersection, all the first vertices from the second historical intersection to the second intersection, and the middle intersection are added to the third array in the order of the first set; the first historical intersection, the middle intersection, the second intersection, and all the first vertices from the second intersection to the first end vertex are added to the fourth array in the order of the first set.
[0109] It is worth mentioning that the root area is obtained by merging multiple sub-areas in the direction corresponding to the cutting line in the current cutting process, and the intermediate intersection is only related to the effective cutting line of the current cutting process, and the historical intermediate intersection does not need to be considered.
[0110] Optionally, in the step of assigning an addressing identifier to each cutting area, the centroid coordinates of each cutting area may be determined; and based on the centroid coordinates of each cutting area, an addressing identifier may be assigned to each cutting area.
[0111] In the embodiment of the present invention, the centroid is also called the geometric center or center point. The geometric center or center point can be calculated by all vertices of the cutting area to obtain the centroid coordinates of the cutting area, and the centroid coordinates are associated with the addressing identifier to assign an addressing identifier to each cutting area. The addressing identifier can be a unique digital number. Figures 2 to 4 As shown in the numbers.
[0112] Optionally, in the step of determining the centroid coordinates of each cutting area, a set of vertices to be calculated corresponding to each cutting area can be obtained, the set of vertices to be calculated including multiple second vertices including a second starting vertex and a second ending vertex of the cutting area, the second vertex being the endpoints of the edge of the cutting area, the multiple second vertices being arranged in a second set order, the second set order being a clockwise order or a counterclockwise order of the cutting area, and each cutting area corresponds to a set of vertices to be calculated; based on the set of vertices to be calculated, determine the directed area between the current second vertex and the remaining second vertices, the directed area being the vector cross product area between the current second vertex and the next second vertex; based on the directed area, determine the weighted cumulative coordinate value of the cutting area and the cumulative directed area of the cutting area; based on the weighted cumulative coordinate value and the cumulative directed area, determine the centroid coordinates of the cutting area.
[0113] In the embodiment of the present invention, the above-mentioned directed area is the "× area" of the vector between the two points. It can be understood that, for two adjacent vertices, the vector between the two points can be the vector from A to B and the vector from B to A. For two non-adjacent vertices, it can also be the vector from A to B and the vector from B to C.
[0114] The process of calculating the centroid of the cutting area is as follows: 1. Initialize variables: xSum, ySum, signedArea, numPoints, where xSum is used to accumulate the weighted x coordinates of the second vertices of the cutting area, ySum is used to accumulate the y coordinates of the second vertices of the cutting area, and signedArea is used to accumulate the directed area of the polygon. numPoints is used to store the number of second vertices of the cutting area. 2. Traverse the second vertices of the cutting area, looping from the first second vertex to the last second vertex, and for each second vertex currentPoint (current second vertex) and the next second vertex nextPoint, calculate the directed area area of the triangle formed between the two points. The above-mentioned directed area is the "× area" of the vector between the two points. The specific calculation formula is: area = currentPoint.x×nextPoint.y-nextPoint.x×currentPoint.y. Accumulate the directed area area, and accumulate the directed area signedArea. After obtaining the directed area, calculate the weighted cumulative x and y coordinates: xSum+ = (currentPoint.x + nextPoint.x) × area, ySum+ = (currentPoint.y + nextPoint.y) × area. When taking the second vertex, a modulo operation is required. The modulo operation (i+1)%numPoints ensures that the next point of the last point is the first point, so that the polygon of the cutting area is closed. 3. Calculate the total directed area of the cutting area: the final signedArea is half of the above total directed area: signedArea× = 0.5 total directed area, 4. Calculate the center of mass coordinates: the x coordinate xSum and y coordinate ySum of the center of mass are xSum+ and ySum+ respectively divided by 6×signedArea to obtain the final center of mass coordinates. The specific calculation is: xSum = xSum+ / 6×signedArea, ySum = ySum+ / 6×signedArea. The centroid coordinates are the final labels, and the addressing mark of the cutting area (the label of the cutting area) is marked on the coordinates.
[0115] By obtaining the set of vertices to be calculated in the cutting area, calculating the directed area and weighted cumulative coordinate values, the centroid coordinates of the cutting area can be accurately calculated.
[0116] After multiple cutting processes, the above-mentioned cutting process is performed on the newly formed root area each time, and the new vertex set formed is continuously updated as the new root area, and finally a refined area collection point required by the business is formed and the cut area is displayed.
[0117] Optionally, in the step of determining the area where the target object is located in the cutting area based on the intersection data between the target detection frame and each cutting area, the reference edge of the target detection frame can be extended toward one end in the target image to obtain a reference straight line of the target detection frame; the number of intersection points between the reference straight line and each cutting area is calculated to determine the area where the target object is located in the cutting area.
[0118] In an embodiment of the present invention, after obtaining the target detection frame, it is compared with the formed cutting area to determine which cutting area the target detection frame overlaps with. The specific judgment method is as follows: 1. Given the coordinates of the four vertices of the target detection frame, traverse each vertex, draw a ray from each vertex, and calculate the number of intersections between this ray and the boundaries of all cutting areas. If the number of intersections is an odd number, it means that the vertex is in the cutting area corresponding to the odd intersection; if it is an even number, the vertex is outside the cutting area corresponding to the even intersection. The calculation of the number of intersections can be as described in the above embodiment. 2. For each cutting area, check whether each vertex of the target detection frame is within the cutting area. If all the vertices of the target detection frame are within a certain cutting area, it is determined that the target detection frame is within the cutting area, and the addressing identifier of the cutting area is obtained. The addressing identifier is transmitted to the robot, and the robot quickly and directly reaches the cutting area to process the business. 3. If a cutting area that is completely contained in all vertices of the target detection box is not found, then a cutting area with three vertices of the target detection box is searched for, and the addressing identifier of the cutting area is obtained. The addressing identifier is transmitted to the robot, and the robot quickly and directly reaches the cutting area to process the business. 4. If a cutting area with three vertices of the target detection box is not found, then a cutting area with two vertices of the target detection box is searched for, and the addressing identifier of the cutting area is obtained. The addressing identifier is transmitted to the robot, and the robot quickly and directly reaches the cutting area to process the business. 5. If a cutting area with two vertices of the target detection box is not found, then a cutting area with one vertex of the target detection box is searched for, and the addressing identifier of the cutting area is obtained. The addressing identifier is transmitted to the robot, and the robot quickly and directly reaches the cutting area to process the business.
[0119] like Figure 5 As shown, an embodiment of the present invention provides an image processing device, the image processing device comprising:
[0120] A first acquisition module 501 is used to acquire multiple cutout areas of a target image, each cutout area corresponds to an addressing identifier;
[0121] A second acquisition module 502 is used to acquire a target detection frame of the target object when a target object is identified in the target picture, wherein the target object is an object that needs to be processed by the robot;
[0122] A first processing module 503 is used to determine the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas;
[0123] The second processing module 504 is used to determine the addressing mark of the area where the target object is located as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
[0124] Optionally, the first acquisition module 501 is also used to determine the root area in the current cutting process in the target picture; in the current cutting process, the root area is subjected to area cutting processing by using the obtained cutting straight line to obtain multiple current cutting areas of the target picture; if the cutting is not completed, the root area for the next cutting process is determined based on the multiple current cutting areas; if the cutting is completed, the multiple cutting areas of the target picture are determined based on the multiple current cutting areas; and each cutting area is assigned an addressing identifier.
[0125] Optionally, the first acquisition module 501 is also used to obtain the cutting line in the current cutting process; perform regional cutting processing on the root area through the cutting line, and determine whether the number of intersections between the cutting line and the root area is two; if the number of intersections between the cutting line and the root area is not two, then the regional cutting process is determined to be invalid, and the cutting line is abandoned; if the number of intersections between the cutting line and the root area is two, then the regional cutting process is determined to be valid, and the cutting line is retained to obtain multiple current sub-cutting areas of the target screen.
[0126] Optionally, the first acquisition module 501 is also used to obtain a first vertex set of the root area, the first vertex set includes multiple first vertices including a first starting vertex and a first ending vertex, the first vertex is the endpoint of the edge of the root area, and the multiple first vertices are arranged in a first set order, and the first set order is a clockwise order or a counterclockwise order of the root area; obtain two intersection points of the cutting straight line and the root area, insert the two intersection points into the vertex set in the first set order to obtain a second vertex set; based on the second vertex set, determine multiple current sub-cutting areas of the target screen.
[0127] Optionally, the first acquisition module 501 is also used to initialize multiple arrays; in the second vertex set, multiple arrays are determined in the order of the first set; based on the multiple arrays, multiple current sub-cutting areas of the target screen are determined, each array corresponds to one current sub-cutting area.
[0128] Optionally, the first acquisition module 501 is further used to determine the centroid coordinates of each cutting area; and assign an addressing identifier to each cutting area based on the centroid coordinates of each cutting area.
[0129] Optionally, the first acquisition module 501 is also used to obtain a set of vertices to be calculated corresponding to each of the cutting areas, the set of vertices to be calculated includes multiple second vertices including a second starting vertex and a second ending vertex of the cutting area, the second vertex is the endpoint of the edge of the cutting area, and the multiple second vertices are arranged in a second set order, the second set order is a clockwise order or a counterclockwise order of the cutting area, and each of the cutting areas corresponds to one set of vertices to be calculated; based on the set of vertices to be calculated, determine the directed area between the current second vertex and the remaining second vertices, the directed area is the vector cross product area between the current second vertex and the next second vertex; based on the directed area, determine the weighted cumulative coordinate value of the cutting area and the cumulative directed area of the cutting area; based on the weighted cumulative coordinate value and the cumulative directed area, determine the centroid coordinates of the cutting area.
[0130] Optionally, the first processing module 503 is also used to extend the reference edge of the target detection frame in one direction in the target image to obtain a reference straight line of the target detection frame; calculate the number of intersection points between the reference straight line and each of the cutting areas, and determine the area where the target object is located in the cutting area.
[0131] like Figure 6 As shown, an embodiment of the present invention further provides an electronic device, including a processor, and the processor can execute any one of the above-mentioned image processing methods.
[0132] Specifically, it includes a processor 601 and a memory 602, and a computer program for executing the image processing method which is stored in the memory 602 and can be run on the processor 601, wherein:
[0133] The processor 601 runs the computer program of the image processing method stored in the memory 602 to perform the following steps:
[0134] Acquire multiple cut areas of the target image, each cut area corresponds to an addressing identifier;
[0135] When a target object is identified in the target picture, a target detection frame of the target object is obtained, wherein the target object is an object that needs to be processed by the robot;
[0136] Determining the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas;
[0137] The addressing mark of the area where the target object is located is determined as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
[0138] Optionally, the acquiring of multiple cut regions of the target picture performed by the processor 601 includes:
[0139] Determining a root area in a current cutting process in the target picture;
[0140] In the current cutting process, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current cutting regions of the target image;
[0141] If the cutting is not completed, determining the root area for the next cutting process based on the multiple current cutting areas;
[0142] If the cutting is completed, determining a plurality of cutting areas of the target image based on the plurality of current cutting areas;
[0143] Each cutting area is assigned an addressing identifier.
[0144] Optionally, in the current sub-cutting process executed by the processor 601, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current sub-cutting regions of the target image, including:
[0145] In the current cutting process, obtaining the cutting straight line;
[0146] Performing a region cutting process on the root region by using the cutting straight line, and determining whether the number of intersections between the cutting straight line and the root region is two;
[0147] If the number of intersections between the cutting line and the root region is not two, the region cutting process is determined to be invalid, and the cutting line is abandoned;
[0148] If there are two intersection points between the cutting line and the root area, the area cutting process is determined to be valid, and the cutting line is retained to obtain a plurality of current sub-cutting areas of the target image.
[0149] Optionally, the processor 601 performs the step of retaining the cutting straight line to obtain a plurality of current sub-cutting areas of the target image, including:
[0150] Obtain a first vertex set of the root region, wherein the first vertex set includes a plurality of first vertices including a first starting vertex and a first ending vertex, the first vertices being endpoints of edges of the root region, and the plurality of first vertices being arranged in a first set order, which is a clockwise order or a counterclockwise order of the root region;
[0151] Obtain two intersection points of the cutting line and the root region, and insert the two intersection points into the vertex set in the order of the first set to obtain a second vertex set;
[0152] Based on the second vertex set, a plurality of current sub-cutting areas of the target picture are determined.
[0153] Optionally, the determining of a plurality of current sub-cutting areas of the target picture based on the second vertex set performed by the processor 601 includes:
[0154] Initialize multiple arrays;
[0155] In the second vertex set, a plurality of arrays are determined in the order of the first set;
[0156] Based on the multiple arrays, multiple current sub-cutting regions of the target picture are determined, and each array corresponds to one current sub-cutting region.
[0157] Optionally, the assigning of an addressing identifier to each cutting area performed by the processor 601 includes:
[0158] Determining the centroid coordinates of each of the cutting areas;
[0159] Based on the centroid coordinates of each cutting area, an addressing identifier is assigned to each cutting area.
[0160] Optionally, the determining of the centroid coordinates of each cutting area performed by the processor 601 includes:
[0161] Obtain a set of vertices to be calculated corresponding to each cutting area, wherein the set of vertices to be calculated includes a plurality of second vertices including a second starting vertex and a second ending vertex of the cutting area, wherein the second vertices are endpoints of edges of the cutting area, and the plurality of second vertices are arranged in a second set order, wherein the second set order is a clockwise order or a counterclockwise order of the cutting area, and each cutting area corresponds to one set of vertices to be calculated;
[0162] Based on the set of vertices to be calculated, determine the directed area between the current second vertex and the remaining second vertices, where the directed area is the vector cross product area between the current second vertex and the next second vertex;
[0163] Based on the directed area, determining the weighted accumulated coordinate value of the cutting area and the accumulated directed area of the cutting area;
[0164] Based on the weighted accumulated coordinate values and the accumulated directed area, the centroid coordinates of the cutting area are determined.
[0165] Optionally, the processor 601 determines, according to the intersection data between the target detection frame and each of the cutting areas, a region where the target object is located in the cutting area, including:
[0166] In the target picture, extending the reference edge of the target detection frame toward one end to obtain a reference straight line of the target detection frame;
[0167] The number of intersections between the reference straight line and each of the cutting areas is calculated, and the area where the target object is located is determined in the cutting area.
[0168] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the image processing method or the application-side image processing method provided by the embodiment of the present invention are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0169] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.
[0170] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.
Claims
1. An image processing method, characterized in that: The method comprises the following steps: Acquire multiple cut areas of the target image, each cut area corresponds to an addressing identifier; When a target object is identified in the target picture, a target detection frame of the target object is obtained, wherein the target object is an object that needs to be processed by the robot; Determining the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas; The addressing mark of the area where the target object is located is determined as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
2. The image processing method according to claim 1, characterized in that: The step of obtaining multiple cut areas of the target image includes: Determining a root area in a current cutting process in the target picture; In the current cutting process, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current cutting regions of the target image; If the cutting is not completed, determining the root area for the next cutting process based on the multiple current cutting areas; If the cutting is completed, then based on the multiple current cutting areas, determining multiple cutting areas of the target image; Each cutting area is assigned an addressing identifier.
3. The image processing method according to claim 2, characterized in that: In the current sub-cutting process, the root region is subjected to region cutting process by using the obtained cutting straight line to obtain a plurality of current sub-cutting regions of the target image, including: In the current cutting process, obtaining the cutting straight line; Performing a region cutting process on the root region by using the cutting straight line, and determining whether the number of intersections between the cutting straight line and the root region is two; If the number of intersections between the cutting line and the root region is not two, the region cutting process is determined to be invalid, and the cutting line is abandoned; If there are two intersection points between the cutting line and the root area, the area cutting process is determined to be valid, and the cutting line is retained to obtain a plurality of current sub-cutting areas of the target image.
4. The image processing method according to claim 3, characterized in that: The step of retaining the cutting straight line to obtain a plurality of current sub-cutting areas of the target image includes: Obtain a first vertex set of the root region, wherein the first vertex set includes a plurality of first vertices including a first starting vertex and a first ending vertex, the first vertices being endpoints of edges of the root region, and the plurality of first vertices being arranged in a first set order, which is a clockwise order or a counterclockwise order of the root region; Obtain two intersection points of the cutting line and the root region, and insert the two intersection points into the vertex set in the order of the first set to obtain a second vertex set; Based on the second vertex set, a plurality of current sub-cutting areas of the target picture are determined.
5. The image processing method according to claim 4, characterized in that: The step of determining a plurality of current sub-cutting areas of the target image based on the second vertex set includes: In the second vertex set, a plurality of arrays are determined in the order of the first set; Based on the multiple arrays, multiple current sub-cutting regions of the target picture are determined, and each array corresponds to one current sub-cutting region.
6. The image processing method according to any one of claims 2 to 5, characterized in that: The step of assigning an addressing identifier to each cutting area comprises: Determining the centroid coordinates of each of the cutting areas; Based on the centroid coordinates of each cutting area, an addressing identifier is assigned to each cutting area.
7. The image processing method according to claim 6, characterized in that: Determining the centroid coordinates of each cutting area includes: Obtain a set of vertices to be calculated corresponding to each cutting area, wherein the set of vertices to be calculated includes a plurality of second vertices including a second starting vertex and a second ending vertex of the cutting area, wherein the second vertices are endpoints of edges of the cutting area, and the plurality of second vertices are arranged in a second set order, wherein the second set order is a clockwise order or a counterclockwise order of the cutting area, and each cutting area corresponds to one set of vertices to be calculated; Based on the set of vertices to be calculated, determine the directed area between the current second vertex and the remaining second vertices, where the directed area is the vector cross product area between the current second vertex and the next second vertex; Based on the directed area, determining the weighted accumulated coordinate value of the cutting area and the accumulated directed area of the cutting area; Based on the weighted accumulated coordinate values and the accumulated directed area, the centroid coordinates of the cutting area are determined.
8. The image processing method according to any one of claims 1 to 7, characterized in that: The step of determining the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas comprises: In the target picture, extending the reference edge of the target detection frame toward one end to obtain a reference straight line of the target detection frame; The number of intersections between the reference straight line and each of the cutting areas is calculated, and the area where the target object is located is determined in the cutting area.
9. An image processing device, characterized in that: The image processing device comprises: A first acquisition module is used to acquire multiple cut areas of the target image, each cut area corresponds to an addressing identifier; A second acquisition module is used to acquire a target detection frame of the target object when a target object is identified in the target picture, wherein the target object is an object that needs to be processed by the robot; A first processing module, configured to determine the area where the target object is located in the cutting area according to the intersection data between the target detection frame and each of the cutting areas; The second processing module is used to determine the addressing mark of the area where the target object is located as the target addressing mark, and the target addressing mark is used to guide the robot to find the area where the target object is located.
10. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the image processing method according to any one of claims 1 to 8 when executing the computer program.
Citation Information
Cited By
Lawn mower global path automatic planning algorithm
CN121613882A
