Spherical or quasi-spherical object positioning method, device and medium based on point cloud information

Abstract

The present invention relates to artificial intelligence technology, and provides a method, device and medium for positioning a spherical or quasi-spherical object based on point cloud information, the method comprising: obtaining an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object comprises a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel; performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; mapping the target pixel area to the depth map of the spherical or quasi-spherical object to determine a target point cloud; performing equal-depth slicing processing on the target point cloud based on a preset target depth value to obtain a set of points to be fitted; performing circle detection on the set of points to be fitted to determine the target fitting circle and the target circle center coordinates corresponding to the target fitting circle. According to the scheme provided by an embodiment of the present invention, accurate positioning can be performed when one or more parts of a spherical or quasi-spherical object are blocked by an obstruction.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, but is not limited to it, and in particular to a method, device and medium for positioning a spherical or quasi-spherical object based on point cloud information. Background Art

[0002] With the advancement and development of science and technology, industrial automation has become the mainstream trend of my country's industry development. Industrial intelligence and automation are the trends of widely using intelligent control and automatic adjustment devices in industrial production to replace manual operation of machines and machine systems for processing and production.

[0003] When an intelligent control device is used to process a spherical or quasi-spherical object, positioning the spherical or quasi-spherical object is an important processing step. Currently, for conventional working environments, an object image is acquired through a visual system that combines a camera with a depth sensor, and then edge detection is performed on the object image to obtain the target boundary. Finally, the Hough transform method is used to fit a circle and determine the center position of the circle, thereby completing the positioning of the object. However, for complex working environments, edge detection is performed on the object image. When one or more parts of the spherical or quasi-spherical object are blocked by an obstruction, the edge detection result cannot determine the edge of the spherical or quasi-spherical object, resulting in the inability to generate an accurate fitting circle and the inability to accurately position the spherical or quasi-spherical object. Summary of the invention

[0004] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0005] Embodiments of the present invention provide a method, device and medium for positioning a spherical or quasi-spherical object based on point cloud information, which can accurately position a spherical or quasi-spherical object when one or more parts of the spherical or quasi-spherical object are blocked by an obstruction.

[0006] In the first aspect, the present invention also provides a method for locating a spherical or quasi-spherical object based on point cloud information, comprising: acquiring an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes multiple pixel points, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel point; performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; mapping the target pixel area to the depth map of the spherical or quasi-spherical object to determine a target point cloud; based on a preset target depth value, performing equal-depth slicing processing on the target point cloud to obtain a set of points to be fitted; performing circle detection on the set of points to be fitted to determine a target fitting circle and the target center coordinates corresponding to the target fitting circle.

[0007] In some embodiments, the set of points to be fitted is a collection of multiple points to be fitted; performing circle detection on the set of points to be fitted to determine a target fitting circle and target circle center coordinates corresponding to the target fitting circle, comprises: determining multiple groups of points to be fitted from the set of points to be fitted, wherein the group of points to be fitted is composed of three points to be fitted that are not on the same straight line; determining fitting circles corresponding to each of the groups of points to be fitted and the corresponding circle center coordinates according to the groups of points to be fitted and a preset three-point circle determination algorithm; performing weighted averaging processing on all the circle center coordinates to obtain target circle center coordinates; and determining a target fitting circle according to the fitting circles and the target circle center coordinates.

[0008] In some embodiments, the weighted averaging of all the circle center coordinates to obtain the target circle center coordinates includes: performing object recognition on the image of the spherical or quasi-spherical object to determine object type information; performing scene recognition on the image of the spherical or quasi-spherical object to determine scene information; determining a weighting coefficient based on the object type information, the scene information and a preset weighting condition; and performing weighted averaging of all the circle center coordinates according to the weighting coefficient to obtain the target circle center coordinates.

[0009] In some embodiments, before the step of performing weighted averaging processing on all the center coordinates of the circle to obtain the target center coordinates of the circle, it also includes: preprocessing each of the center coordinates of the circle; wherein the preprocessing includes at least one of the following: filtering each of the center coordinates of the circle according to preset filtering conditions; performing abnormal elimination processing on each of the center coordinates of the circle.

[0010] In some embodiments, the abnormal elimination processing of each of the center coordinates includes: calculating the coordinate standard deviation and coordinate mean of all the center coordinates; determining abnormal coordinates from all the center coordinates based on the Laida criterion, the coordinate standard deviation and the coordinate mean; and eliminating the abnormal coordinates from all the center coordinates.

[0011] In some embodiments, determining a plurality of groups of points to be fitted from the set of points to be fitted comprises: sorting all the points to be fitted in the set of points to be fitted to determine a sequence of points to be fitted; sampling the sequence of points to be fitted multiple times according to a preset sampling interval, or randomly sampling the sequence of points to be fitted multiple times to determine a plurality of groups of points to be fitted.

[0012] In some embodiments, the semantic segmentation of the image of the spherical or quasi-spherical object to obtain the target pixel area includes: determining the red channel value, green channel value and blue channel value of each pixel; determining the color difference of the pixel according to the red channel value, green channel value and blue channel data; screening the pixels according to the preset difference threshold and the color difference to determine the target pixel; and obtaining the target pixel area based on the target pixel.

[0013] In some embodiments, the semantic segmentation of the image of the spherical or quasi-spherical object to obtain the target pixel area includes: inputting the image of the spherical or quasi-spherical object into a trained spherical or quasi-spherical object segmentation model to obtain the target pixel area; wherein the spherical or quasi-spherical object segmentation model is trained by the following steps: acquiring a spherical or quasi-spherical object training image and a spherical or quasi-spherical object training label image; inputting the spherical or quasi-spherical object training image into a preset semantic segmentation model to obtain a predicted area; determining a cross entropy loss function based on the spherical or quasi-spherical object training label image and the predicted area; and updating the model parameters of the semantic segmentation model based on the cross entropy loss function to obtain a spherical or quasi-spherical object segmentation model.

[0014] In some embodiments, performing circle detection on the set of points to be fitted to determine a target fitting circle and target circle center coordinates corresponding to the target fitting circle includes: performing fitting processing on the set of points to be fitted according to a least squares method, or performing circle detection on the set of points to be fitted according to a Hough transform method to determine a target fitting circle; determining the target circle center coordinates according to the target fitting circle.

[0015] In a second aspect, the present invention also provides a spherical or quasi-spherical object positioning device based on point cloud information, comprising: an acquisition unit, used to acquire an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes multiple pixels, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel; a segmentation unit, used to perform semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; a point cloud determination unit, used to map the target pixel area to the depth map of the spherical or quasi-spherical object to determine the target point cloud; a point set determination unit, used to perform equal-depth slicing processing on the target point cloud based on a preset target depth value to obtain a point set to be fitted; a positioning unit, used to perform circle detection on the point set to be fitted, and determine the target fitting circle and the target circle center coordinates corresponding to the target fitting circle.

[0016] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for locating a spherical or quasi-spherical object based on point cloud information as described in the first aspect above is implemented.

[0017] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the method for locating a spherical or quasi-spherical object based on point cloud information as described in the first aspect above.

[0018] An embodiment of the present invention includes: acquiring an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel; performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; mapping the target pixel area to the depth map of the spherical or quasi-spherical object to determine a target point cloud; performing equal-depth slicing processing on the target point cloud based on a preset target depth value to obtain a set of points to be fitted; performing circle detection on the set of points to be fitted to determine a target fitting circle and the target circle center coordinates corresponding to the target fitting circle. According to the solution provided by the embodiment of the present invention, by performing semantic segmentation on the image of a spherical or quasi-spherical object, the obtained target pixel area is mapped to a depth map of the spherical or quasi-spherical object to obtain a target point cloud, and through depth slicing processing, a set of points to be fitted is obtained from the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by an occluder.

[0019] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0021] Figure 1 A schematic diagram of an implementation environment provided for an embodiment of the present invention;

[0022] Figure 2is a flow chart of a method for locating a spherical or quasi-spherical object based on point cloud information provided by an embodiment of the present invention;

[0023] Figure 3 is a flow chart of determining a target fitting circle and target circle center coordinates provided by another embodiment of the present invention;

[0024] Figure 4 is a flow chart for determining the coordinates of the center of a target circle provided by another embodiment of the present invention;

[0025] Figure 5 is a flow chart of a circle center coordinate preprocessing provided by another embodiment of the present invention;

[0026] Figure 6 is a flow chart of eliminating abnormal coordinates provided by another embodiment of the present invention;

[0027] Figure 7 is a flow chart for determining a group of points to be fitted provided by another embodiment of the present invention;

[0028] Figure 8 is a flow chart of determining a target pixel area provided by another embodiment of the present invention;

[0029] Fig. 9 is another flow chart for determining a target pixel area provided by another embodiment of the present invention;

[0030] Fig.10 is another flow chart for determining the target fitting circle and the target center coordinates provided by another embodiment of the present invention;

[0031] Fig.11 is a schematic diagram of an image provided by another embodiment of the present invention;

[0032] Fig.12 is a schematic diagram of a target pixel area provided by another embodiment of the present invention;

[0033] Fig.13 is a schematic diagram of a target point cloud provided by another embodiment of the present invention;

[0034] Fig.14 is a schematic diagram of a set of points to be fitted provided by another embodiment of the present invention;

[0035] Fig.15 is a structural diagram of a spherical or quasi-spherical object positioning device based on point cloud information provided by another embodiment of the present invention;

[0036] Fig.16 It is a structural diagram of an electronic device provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0038] In the description of the present invention, “several” means one or more, “more” means more than two, “greater than”, “less than”, “exceed”, etc. are understood as not including the number, and “above”, “below”, “within”, etc. are understood as including the number.

[0039] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims or the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0040] Before further describing the embodiments of the present invention in detail, the nouns and terms involved in the embodiments of the present invention are described. The nouns and terms involved in the embodiments of the present invention are applicable to the following interpretations:

[0041] Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is also the study of the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making. The basic technologies of artificial intelligence generally include sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics and other technologies. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology and machine learning / deep learning. With the research and progress of artificial intelligence technology, artificial intelligence has been studied and applied in many fields, such as common smart homes, smart customer service, virtual assistants, smart speakers, smart marketing, unmanned driving, automatic driving, robots, smart medical care, smart cities, etc. It is believed that with the development of technology, artificial intelligence will be applied in more fields and play an increasingly important role.

[0042] Machine Learning (ML) is a multi-disciplinary interdisciplinary subject that involves probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in studying how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent. Its applications are spread across all areas of artificial intelligence. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning and other technologies.

[0043] Spherical and quasi-spherical objects refer to standard spherical objects, ellipsoids and irregular spheres, such as apples, pears, peaches, oranges, tangerines, coconuts and other fruits.

[0044] Equal depth slicing, extracting a certain point cloud information with the same depth value from the depth point cloud information of the target object to form a point cloud set.

[0045] Structured light is a system structure composed of a projector and a camera. The projector projects specific light information onto the surface of an object and the background. The camera collects the information and calculates the position and depth of the object based on the changes in the light signal caused by the object, thereby restoring the entire three-dimensional space.

[0046] Time of flight (TOF) can be broadly understood as a technology that measures the time it takes for an object, particle or wave to fly a certain distance in a fixed medium (the medium / distance / time are all known or measurable) to further understand certain properties of ions or media.

[0047] The Laida criterion means that we first assume that a set of test data contains only random errors, calculate and process them to obtain the standard deviation, and determine an interval according to a certain probability. It is believed that any error exceeding this interval is not a random error but a gross error, and the data containing this error should be eliminated.

[0048] The least squares method is a mathematical tool that is widely used in many disciplines of data processing, such as error estimation, uncertainty, system identification, prediction, and forecast.

[0049] Hough transform is a feature extraction method that is widely used in image analysis, computer vision, and digital image processing. Hough transform can not only identify straight lines, but also any shapes, the most common of which are circles and ellipses.

[0050] Cross Entropy is mainly used to measure the difference information between two probability distributions.

[0051] When an intelligent control device is used to process a spherical or quasi-spherical object, positioning the spherical or quasi-spherical object is an important processing step. Currently, for conventional working environments, an object image is acquired through a visual system that combines a camera with a depth sensor, and then edge detection is performed on the object image to obtain the target boundary. Finally, the Hough transform method is used to fit a circle and determine the center position of the circle, thereby completing the positioning of the object. However, for complex working environments, edge detection is performed on the object image. When one or more parts of the spherical or quasi-spherical object are blocked by an obstruction, the edge detection result cannot determine the edge of the spherical or quasi-spherical object, resulting in the inability to generate an accurate fitting circle and the inability to accurately position the spherical or quasi-spherical object.

[0052] In view of the problem that a spherical or quasi-spherical object cannot be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by an obstruction, the present invention provides a spherical or quasi-spherical object positioning method, device and medium based on point cloud information, the method comprising: acquiring an image and a depth map of the spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object comprises a plurality of pixel points, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel point; performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; mapping the target pixel area to the depth map of the spherical or quasi-spherical object to determine a target point cloud; performing equal-depth slicing processing on the target point cloud based on a preset target depth value to obtain a set of points to be fitted; performing circle detection on the set of points to be fitted to determine a target fitting circle and the target circle center coordinates corresponding to the target fitting circle. According to the solution provided by the embodiment of the present invention, by performing semantic segmentation on the image of a spherical or quasi-spherical object, the obtained target pixel area is mapped to a depth map of the spherical or quasi-spherical object to obtain a target point cloud, and through depth slicing processing, a set of points to be fitted is obtained from the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by an occluder.

[0053] A method for positioning a spherical or quasi-spherical object based on point cloud information provided in an embodiment of the present invention can be applied in a terminal or a server, and can also be software running in a terminal or a server.

[0054] Reference Figure 1 , Figure 1A schematic diagram of an implementation environment provided for an embodiment of the present invention, the implementation environment includes a terminal 101 and a server 102 , wherein the terminal 101 and the server 102 are connected via a communication network 103 .

[0055] Server 102 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), as well as big data and artificial intelligence platforms.

[0056] In addition, server 102 may also be a node server in the blockchain network.

[0057] The terminal 101 may be a personal computer (PC), a mobile phone, a smart phone, a personal digital assistant (PDA), a wearable device, a handheld computer (PPC), a tablet computer, a shooting device with a shooting function, a collection device, etc., but is not limited thereto. The terminal 101 and the server 102 may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present invention.

[0058] It should be noted that the intelligent control device can be connected to the terminal 101 or the server 102 via a network, so that the spherical or quasi-spherical object positioning method provides functional support for the intelligent control device.

[0059] based on Figure 1 The implementation environment, such as Figure 2 As shown, Figure 2 The flowchart of a method for locating a spherical or quasi-spherical object based on point cloud information provided by an embodiment of the present invention includes but is not limited to the following steps:

[0060] Step S210, acquiring an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate a depth value of each pixel;

[0061] Step S220, performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area;

[0062] Step S230, mapping the target pixel area to a depth map of a spherical or quasi-spherical object to determine a target point cloud;

[0063] Step S240, based on a preset target depth value, performing equal depth slicing processing on the target point cloud to obtain a point set to be fitted;

[0064] Step S250, performing circle detection on the set of points to be fitted, and determining the target fitting circle and the target circle center coordinates corresponding to the target fitting circle.

[0065] It can be understood that when one or more parts of a spherical or quasi-spherical object are blocked by an occluder, relative to edge detection of the object image, by acquiring the image and depth map of the spherical or quasi-spherical object, and then performing semantic segmentation on the image of the spherical or quasi-spherical object, the target pixel area of ​​the area where the spherical or quasi-spherical object is located is obtained, and then the target pixel area is mapped to the depth map of the spherical or quasi-spherical object, and the target point cloud containing point cloud information is determined. The point cloud information refers to the position information and depth information of each point in the point cloud. Through multiple experiments, the target depth value of the spherical or quasi-spherical object is determined, and the points in the target point cloud whose depth values ​​meet the target depth values ​​are taken as points to be fitted. , thereby forming a set of points to be fitted, and then performing circle detection on the set of points to be fitted, and then determining the target fitting circle and the target circle center coordinates corresponding to the target fitting circle; based on this, by performing semantic segmentation on the image of the spherical or quasi-spherical object, the obtained target pixel area is mapped to the depth map of the spherical or quasi-spherical object to obtain the target point cloud, and through depth slicing processing, the set of points to be fitted is obtained from the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by an occluder.

[0066] It should be noted that cameras for acquiring images and depth maps of spherical or quasi-spherical objects include but are not limited to: structured light depth cameras, binocular depth cameras, and optical time-of-flight depth cameras; images of spherical or quasi-spherical objects include but are not limited to: color images, pseudo-color images, grayscale images, and black-and-white images.

[0067] In specific practice, the steps for determining the target depth value of a spherical or quasi-spherical object are as follows:

[0068] Step 210, according to different types of spherical and quasi-spherical objects, setting the depth selection ratio λ of the equal-depth slices of the spherical and quasi-spherical objects within a threshold range of 0.5 to 1.0, for example, λ is set to 0.8;

[0069] Step 220, determine the depth range of the target point cloud [Deep min ,Deep max ], for example, Deep min =720, Deep max =755mm;

[0070] Step 230, according to the depth slice depth value calculation formula, the depth slice depth value is calculated by the depth selection ratio λ and the depth range. The depth slice depth value calculation formula is as follows:

[0071] (Deep max -Deep min )*λ+Deep min ,

[0072] Among them, Deep min is the minimum value of the depth range, Deep max is the maximum value of the depth range, λ is the depth selection ratio;

[0073] For example, when λ is set to 0.8, Deep min 720mm, Deep max It is 755mm. According to the calculation formula of the depth slice depth value, we get (755mm-720mm)*0.8+720mm=748mm. Then, the points on the equal depth slice with a depth value of 748mm in the target point cloud are taken as the points to be fitted to form a set of points to be fitted.

[0074] In addition, refer to Figure 3 In one embodiment, the set of points to be fitted is a set of multiple points to be fitted; Figure 2 Step S250 in the illustrated embodiment includes but is not limited to the following steps:

[0075] Step S310, determining a plurality of to-be-fitted point groups from the to-be-fitted point set, wherein the to-be-fitted point group is composed of three to-be-fitted points that are not on the same straight line;

[0076] Step S320, determining the fitting circles corresponding to the to-be-fitted point groups and the corresponding center coordinates according to the to-be-fitted point groups and the preset three-point circle determination algorithm;

[0077] Step S330, performing weighted average processing on all circle center coordinates to obtain target circle center coordinates;

[0078] Step S340, determining the target fitting circle according to the fitting circle and the target circle center coordinates.

[0079] It can be understood that the set of points to be fitted includes multiple points to be fitted, and three points to be fitted that are not on the same straight line are taken as a group. Multiple groups of points to be fitted are determined from the set of points to be fitted, and the three-point circle determination algorithm is used to calculate each group of points to be fitted to obtain multiple fitting circles and the center coordinates of each fitting circle, and the average value of all the center coordinates is calculated to obtain the target center coordinates, and the radii of all fitting circles are weighted averaged to obtain the target radius, and the average fitting circle obtained by the target radius and the target center coordinates is used. When the trajectory of the average fitting circle is within the trajectory of the maximum fitting circle and the trajectory of the minimum fitting circle, the average fitting circle is used as the target fitting circle.

[0080] It should be noted that, assuming that the three points to be fitted are P1(x1, y1), P2(x2, y2), and P3(x3, y3), the equation of the circle determined by the three points to be fitted is:

[0081] (x-x0) 2 +(y-y0) 2 =r 2 ,

[0082] Among them, the center of the circle is O(x0,y0), the radius is r, the point (x,y) is a point on the circle, and the relationship between the three points to be fitted and the center and radius is:

[0083]

[0084] It can be solved:

[0085]

[0086] Then we get the coordinates of the center of the circle O(x0,y0).

[0087] In specific practice, in the process of determining the group of points to be fitted, after three points to be fitted are selected, the three points to be fitted are judged. When the three points to be fitted are on the same straight line, three points to be fitted need to be reselected.

[0088] In addition, refer to Figure 4 , in one embodiment, Figure 3 Step S330 in the illustrated embodiment includes but is not limited to the following steps:

[0089] Step S410, performing object recognition on the image of the spherical or quasi-spherical object to determine the object type information;

[0090] Step S420, performing scene recognition on the image of the spherical or quasi-spherical object to determine scene information;

[0091] Step S430, determining a weighting coefficient according to the object type information, the scene information and a preset weighting condition;

[0092] Step S440: Perform weighted averaging processing on all circle center coordinates according to the weighting coefficient to obtain the target circle center coordinates.

[0093] It can be understood that the averaging process preferably adopts weighted averaging process. Object recognition and scene recognition are both target detection methods. Different weighting coefficients are set for different objects and different scenes. The positions of the centers of different objects are not exactly the same, and the positions of the centers of objects in different scenes may also change. By setting the weighting coefficients, the weighting coefficients of points farther from the center of the object are smaller, and the weighting coefficients of points closer to the center of the object are larger, so that the calculated coordinates of the center of the target circle are more accurate.

[0094] It is worth noting that the methods of object recognition and scene recognition are technologies well known to those skilled in the art and will not be elaborated here.

[0095] In addition, refer to Figure 5 , in one embodiment, Figure 3 Before step S330 in the illustrated embodiment, the following steps are also included but not limited to:

[0096] Step S510, pre-processing each circle center coordinate;

[0097] Wherein, the preprocessing includes at least one of the following:

[0098] According to the preset filtering conditions, filter the coordinates of each circle center;

[0099] Perform abnormal elimination on each circle center coordinate.

[0100] It can be understood that, through preprocessing, abnormal circle center coordinates can be removed, thereby ensuring the accuracy of the target fitting circle and the target circle center coordinates.

[0101] It should be noted that the user sets the size threshold of the fitted circle according to the size of the spherical or quasi-spherical object. The center coordinates of the fitted circle that do not meet the size threshold are abnormal center coordinates. The filtering condition is set to remove the abnormal center coordinates; the abnormal elimination process can also eliminate abnormal center coordinates.

[0102] In addition, refer to Figure 6 , in one embodiment, Figure 5 The step of performing abnormal elimination processing on each circle center coordinate in step S510 in the illustrated embodiment includes but is not limited to the following steps:

[0103] Step S610, calculating the coordinate standard deviation and coordinate mean of all circle center coordinates;

[0104] Step S620, determining abnormal coordinates from all circle center coordinates based on the Laida criterion, coordinate standard deviation and coordinate mean;

[0105] Step S630, eliminating abnormal coordinates from all circle center coordinates.

[0106] It should be noted that when using the Laida criterion for anomaly processing, it is necessary to first calculate the standard deviation and mean, so the coordinate standard deviation and coordinate mean are calculated to determine the abnormal coordinates.

[0107] It is worth noting that the processing method based on the Laida criterion is a technology well known to those skilled in the art and will not be elaborated here.

[0108] like Figure 7 As shown, in one embodiment, Figure 3 Step S310 in the illustrated embodiment includes but is not limited to the following steps:

[0109] Step S710, sorting all the points to be fitted in the set of points to be fitted to determine a sequence of the points to be fitted;

[0110] Step S720: sampling the sequence of points to be fitted multiple times according to a preset sampling interval, or performing random sampling on the sequence of points to be fitted multiple times to determine multiple groups of points to be fitted.

[0111] It should be noted that the sorting method includes but is not limited to: when generating a set of points to be fitted, the points to be fitted in the set of points to be fitted are determined in sequence, and the points to be fitted are sorted according to the order in which the points to be fitted are determined, so as to determine the sequence of the points to be fitted, which can simplify the operation steps and thus improve efficiency.

[0112] It is worth noting that the random sampling method is a technique well known to those skilled in the art and will not be elaborated here.

[0113] In specific practice, a reasonable number of sampling times is set according to the irregularity of the spheroid and the image resolution. The preferred number of sampling times is between 20 and 1000. The more sampling times, the more accurate the positioning result, and the fewer sampling times, the shorter the calculation time. It should be noted that the number of sampling times selected by the user is determined according to the actual situation and is not limited here.

[0114] like Figure 8 As shown, in one embodiment, Figure 2 Step S220 in the illustrated embodiment includes but is not limited to the following steps:

[0115] Step S810, determining the red channel value, green channel value and blue channel value of each pixel;

[0116] Step S820, determining the color difference of the pixel point according to the red channel value, the green channel value and the blue channel data;

[0117] Step S830, screening the pixels according to the preset difference threshold and color difference to determine the target pixel;

[0118] Step S840, obtaining a target pixel area according to the target pixel point.

[0119] It can be understood that each pixel has three color channels, namely, a red channel, a green channel and a blue channel, and the color difference of the pixel refers to the difference between the red channel value and the green channel value, or the difference between the red channel value and the blue channel value, or the difference between the green channel value and the blue channel value; in an image of a spherical or quasi-spherical object, when the difference between the color of the spherical or quasi-spherical object and the color of the image background is large, and when the difference between the color of the spherical or quasi-spherical object and the color of the occluding object is large, by calculating the color difference of each pixel and combining with a preset difference threshold, the target pixel area where the spherical or quasi-spherical object is located can be accurately obtained.

[0120] It is worth noting that determining the values ​​of each color channel of a pixel is a technique well known to those skilled in the art and will not be elaborated herein.

[0121] In specific practice, when the color of a spherical or quasi-spherical object is red, the color of the image background is white, and the color of the occluder is green, the difference between the red channel value and the green channel value is taken as the target color difference. The target color difference of each pixel is calculated, and the difference threshold is set to a larger value. It is determined whether the target color difference of the pixel is greater than the difference threshold. If so, it means that the pixel is in the area where the spherical or quasi-spherical object is located, and the pixel is taken as the target pixel; otherwise, the next pixel is determined; after all pixels are determined, all target pixels are screened to obtain the target pixel area where the spherical or quasi-spherical object is located.

[0122] like Fig. 9 As shown, in one embodiment, Figure 2 Step S220 in the illustrated embodiment includes but is not limited to the following steps:

[0123] Step S910, inputting the image of the spherical or quasi-spherical object into the trained spherical or quasi-spherical object segmentation model to obtain a target pixel area;

[0124] The spherical or quasi-spherical object segmentation model is trained through the following steps:

[0125] Step 910, obtaining a spherical or quasi-spherical object training image and a spherical or quasi-spherical object training label image;

[0126] Step 920, inputting the spherical or quasi-spherical object training image into a preset semantic segmentation model to obtain a predicted area;

[0127] Step 930, determining a cross entropy loss function according to the spherical or quasi-spherical object training label image and the prediction area;

[0128] Step 940: based on the cross entropy loss function, update the model parameters of the semantic segmentation model to obtain a spherical or quasi-spherical object segmentation model.

[0129] It can be understood that the cross entropy loss function is determined by training label images and prediction areas of spherical or quasi-spherical objects, thereby determining the cross entropy loss, and then the cross entropy loss is used to update the model parameters of the semantic segmentation model, so that the spherical or quasi-spherical object segmentation model can improve the segmentation accuracy, and then the spherical or quasi-spherical object segmentation model can accurately obtain the target pixel area where the spherical or quasi-spherical object is located.

[0130] In specific practice, the semantic segmentation model uses the Unet segmentation network.

[0131] It is worth noting that the calculation method of the cross entropy loss function and the method of obtaining the Unet segmentation network are technologies well known to those skilled in the art and will not be elaborated here.

[0132] like Fig.10 As shown, in one embodiment, Figure 2 Step S250 in the illustrated embodiment includes but is not limited to the following steps:

[0133] Step S1010, performing fitting processing on the fitting point set according to the least square method, or performing circle detection on the fitting point set according to the Hough transform method to determine the target fitting circle;

[0134] Step S1020, determining the target circle center coordinates according to the target fitting circle.

[0135] It can be understood that the method of performing circle detection on the fitting point set also includes but is not limited to: least squares method and Hough transform method; the least squares method can be used for curve fitting, which can improve the efficiency of generating the target fitting circle; circle detection through Hough transform method has the advantage of strong anti-interference ability.

[0136] It is worth noting that the calculation method of the least square method or the Hough transform method is a technology well known to those skilled in the art and will not be elaborated here.

[0137] In addition, refer to Figures 11 to 14 , Fig.11 is a schematic diagram of an image provided by another embodiment of the present invention, Fig.12is a schematic diagram of a target pixel area provided by another embodiment of the present invention, Fig.13 is a schematic diagram of a target point cloud provided by another embodiment of the present invention, Fig.14 It is a schematic diagram of a set of points to be fitted provided by another embodiment of the present invention.

[0138] It should be noted that with the rapid development of global informatization, digital agriculture has developed into a new trend in modern agriculture. Using machine vision technology to accurately obtain the number of fruits on fruit trees is the key to fruit tree yield measurement and fine orchard management. Traditional fruit yield estimation is based on historical data, weather conditions and manual counting. This method not only has large errors, but is also time-consuming and labor-intensive, and can no longer meet the needs of modern agricultural development; with the development of modern information technology, the yield measurement method based on machine vision has shown the potential to replace the manual yield measurement method; at present, in the process of automated harvesting of agricultural products, it is necessary to obtain the spatial position of the fruit picking point. For spherical or quasi-spherical fruits, the fruit image is first obtained using a visual system combining a camera and a depth sensor, and then the fruit image is edge detected to obtain the target boundary. Finally, the Hough transform method is used to fit the circle and determine the center position of the circle, so as to obtain the fruit picking point. However, in a complex growth environment, after the spherical or quasi-spherical fruits are blocked by obstructions such as leaves, trunks, and branches, the edge detection result cannot effectively generate a fitting circle, resulting in the inability to accurately locate the spherical or quasi-spherical fruits.

[0139] It can be understood that the spherical or quasi-spherical object positioning method based on point cloud information in the embodiment of the present invention can be applied to digital agriculture, such as an automatic picking system for spherical or quasi-spherical fruits. Fig.11 The image containing spherical or quasi-spherical fruits shown in the figure is semantically segmented to obtain the following Fig.12 The target pixel area shown in FIG. 1 is mapped to the depth map to obtain the target pixel area as shown in FIG. 1 . Fig.13 The target point cloud shown in FIG. 1 is determined by the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical fruit. Fig.14 For example, the target depth value is 748 mm. Through equal depth slicing, all points with a depth value of 748 mm in the point cloud are extracted to form a set of points to be fitted, and placed on the plane coordinate system composed of the X-axis and Y-axis. Then, the fitting circle and the center coordinates are determined through circle detection, as shown in Fig.14 As shown, the circular points are the points to be fitted, the dotted circle is the fitting circle, the square points are the center of the circle, and the triangular points are the points in the group of points to be fitted. The target circle center coordinates are determined by preprocessing, and then the target fitting circle is determined. This allows the spherical or quasi-spherical fruit to be accurately positioned when one or more parts of the spherical or quasi-spherical fruit are blocked by obstructions, thereby ensuring the normal operation of the automated picking system for spherical or quasi-spherical fruits.

[0140] In addition, refer to Fig.15 The present invention also provides a spherical or quasi-spherical object positioning device 1500 based on point cloud information, comprising:

[0141] An acquisition unit 1510 is used to acquire an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate a depth value of each pixel;

[0142] The segmentation unit 1520 is used to perform semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area;

[0143] A point cloud determination unit 1530 is used to map the target pixel area to a depth map of a spherical or quasi-spherical object to determine a target point cloud;

[0144] The point set determination unit 1540 is used to perform equal-depth slicing processing on the target point cloud based on a preset target depth value to obtain a point set to be fitted;

[0145] The positioning unit 1550 is used to perform circle detection on the set of fitting points to be fitted, and determine the target fitting circle and the target circle center coordinates corresponding to the target fitting circle.

[0146] It can be understood that the specific implementation of the spherical or quasi-spherical object positioning device 1500 based on point cloud information is basically the same as the specific implementation example of the spherical or quasi-spherical object positioning method based on point cloud information mentioned above, and will not be repeated here; based on this, by performing semantic segmentation on the image of the spherical or quasi-spherical object, the obtained target pixel area is mapped to the depth map of the spherical or quasi-spherical object to obtain the target point cloud, and through depth slicing processing, the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object are used to obtain the set of points to be fitted, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by occluders.

[0147] In addition, refer to Fig.16 An embodiment of the present invention further provides an electronic device, the electronic device 1600 comprising: a memory 1610, a processor 1620, and a computer program stored in the memory 1610 and executable on the processor 1620.

[0148] The processor 1620 and the memory 1610 may be connected via a bus or other means.

[0149] The non-transient software program and instructions required to implement the spherical or quasi-spherical object positioning method based on point cloud information in the above embodiment are stored in the memory 1610. When executed by the processor 1620, the spherical or quasi-spherical object positioning method based on point cloud information in the above embodiment is executed, for example, the above described Figure 2 Steps S210 to S250 of the method, Figure 3 Steps S310 to S340 of the method, Figure 4 Steps S410 to S440 of the method, Figure 5 Step S510 of the method, Figure 6 Steps S610 to S630 of the method, Figure 7 Steps S710 to S720 of the method, Figure 8 Steps S810 to S840 of the method, Fig. 9 In the method step S910, Fig.10 In the method steps S1010 to S1020, an image and a depth map of a spherical or quasi-spherical object are obtained, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel; semantic segmentation is performed on the image of the spherical or quasi-spherical object to obtain a target pixel area; the target pixel area is mapped to the depth map of the spherical or quasi-spherical object to determine a target point cloud; based on a preset target depth value, the target point cloud is sliced ​​to obtain a set of points to be fitted; circle detection is performed on the set of points to be fitted to determine a target fitting circle and the target center coordinates corresponding to the target fitting circle. Based on this, by performing semantic segmentation on the image of spherical or quasi-spherical objects, the obtained target pixel area is mapped to the depth map of the spherical or quasi-spherical object to obtain the target point cloud, and through depth slicing processing, the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object are used to obtain the set of points to be fitted, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by occluders.

[0150] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0151] In addition, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions. The computer-executable instructions are executed by a processor or a controller, for example, by a processor in the above electronic device embodiment, so that the above processor can execute the spherical or quasi-spherical object positioning method based on point cloud information in the above embodiment, for example, executing the above described Figure 2 Steps S210 to S250 of the method, Figure 3 Steps S310 to S340 of the method, Figure 4 Steps S410 to S440 of the method, Figure 5 Step S510 of the method, Figure 6 Steps S610 to S630 of the method, Figure 7 Steps S710 to S720 of the method, Figure 8 Steps S810 to S840 of the method, Fig. 9 In the method step S910, Fig.10 In the method steps S1010 to S1020, an image and a depth map of a spherical or quasi-spherical object are obtained, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate the depth value of each pixel; semantic segmentation is performed on the image of the spherical or quasi-spherical object to obtain a target pixel area; the target pixel area is mapped to the depth map of the spherical or quasi-spherical object to determine a target point cloud; based on a preset target depth value, the target point cloud is sliced ​​to obtain a set of points to be fitted; circle detection is performed on the set of points to be fitted to determine a target fitting circle and the target center coordinates corresponding to the target fitting circle. Based on this, by performing semantic segmentation on the image of spherical or quasi-spherical objects, the obtained target pixel area is mapped to the depth map of the spherical or quasi-spherical object to obtain the target point cloud, and through depth slicing processing, the point cloud information of the target point cloud and the target depth value of the spherical or quasi-spherical object are used to obtain the set of points to be fitted, and then the target fitting circle and the target circle center coordinates are determined through circle detection, so that the spherical or quasi-spherical object can be accurately positioned when one or more parts of the spherical or quasi-spherical object are blocked by occluders.

[0152] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0153] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present invention. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A method for locating spherical or quasi-spherical objects based on point cloud information, characterized in that: include: Acquire an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate a depth value of each of the pixels; Performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; Mapping the target pixel area to a depth map of the spherical or quasi-spherical object to determine a target point cloud; Based on the preset target depth value, the target point cloud is sliced ​​with equal depth to obtain a set of points to be fitted; wherein the steps for determining the preset target depth value are as follows: According to different types of the spherical or quasi-spherical objects, setting a depth selection ratio of equal-depth slices of the spherical or quasi-spherical objects within a preset threshold range; Determining a depth range of the target point cloud; The depth slice depth value is calculated by the depth selection ratio and the depth range; Perform circle detection on the set of points to be fitted, and determine a target fitting circle and target circle center coordinates corresponding to the target fitting circle.

2. The method according to claim 1, characterized in that: The set of points to be fitted is a collection of multiple points to be fitted; The performing circle detection on the set of points to be fitted to determine a target fitting circle and target circle center coordinates corresponding to the target fitting circle includes: Determine a plurality of to-be-fitted point groups from the to-be-fitted point set, wherein the to-be-fitted point group is composed of three to-be-fitted points that are not on the same straight line; Determine the fitting circles corresponding to the respective groups of points to be fitted and the corresponding coordinates of the center of the circles according to the groups of points to be fitted and a preset three-point circle determination algorithm; Performing weighted average processing on all the circle center coordinates to obtain the target circle center coordinates; A target fitting circle is determined according to the fitting circle and the target circle center coordinates.

3. The method according to claim 2, characterized in that The weighted average processing of all the circle center coordinates is performed to obtain the target circle center coordinates, including: Performing object recognition on the image of the spherical or quasi-spherical object to determine object type information; Performing scene recognition on the image of the spherical or quasi-spherical object to determine scene information; Determining a weighting coefficient according to the object type information, the scene information and a preset weighting condition; According to the weighting coefficient, all the circle center coordinates are weighted averaged to obtain the target circle center coordinates.

4. The method according to claim 2, characterized in that: Before the step of performing weighted average processing on all the circle center coordinates to obtain the target circle center coordinates, the method further includes: Preprocessing each of the circle center coordinates; Wherein, the preprocessing includes at least one of the following: According to the preset filtering conditions, each of the circle center coordinates is filtered; An abnormal elimination process is performed on each of the circle center coordinates.

5. The method according to claim 4, characterized in that The abnormal elimination process for each of the circle center coordinates includes: Calculate the coordinate standard deviation and coordinate mean of all the circle center coordinates; Based on the Laida criterion, the coordinate standard deviation and the coordinate mean, determining abnormal coordinates from all the circle center coordinates; Eliminate the abnormal coordinates among all the circle center coordinates.

6. The method according to claim 2, characterized in that The step of determining a plurality of groups of points to be fitted from the set of points to be fitted comprises: Sorting all the points to be fitted in the set of points to be fitted to determine a sequence of points to be fitted; The sequence of points to be fitted is sampled multiple times according to a preset sampling interval, or the sequence of points to be fitted is randomly sampled multiple times to determine multiple groups of points to be fitted.

7. The method according to claim 1, characterized in that The performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area includes: Determine the red channel value, green channel value and blue channel value of each pixel; Determine the color difference of the pixel point according to the red channel value, the green channel value and the blue channel data; The pixel points are screened according to a preset difference threshold and the color difference to determine a target pixel point; According to the target pixel point, a target pixel area is obtained.

8. The method according to claim 1, characterized in that The performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area includes: Inputting the image of the spherical or quasi-spherical object into a trained spherical or quasi-spherical object segmentation model to obtain a target pixel region; The spherical or quasi-spherical object segmentation model is trained by the following steps: Obtaining a spherical or quasi-spherical object training image and a spherical or quasi-spherical object training label image; Inputting the spherical or quasi-spherical object training image into a preset semantic segmentation model to obtain a predicted area; Determining a cross entropy loss function according to the spherical or quasi-spherical object training label image and the prediction area; Based on the cross entropy loss function, the model parameters of the semantic segmentation model are updated to obtain a spherical or quasi-spherical object segmentation model.

9. The method according to claim 1, characterized in that: The performing circle detection on the set of points to be fitted to determine a target fitting circle and target circle center coordinates corresponding to the target fitting circle includes: Performing fitting processing on the set of points to be fitted according to the least squares method, or performing circle detection on the set of points to be fitted according to the Hough transform method to determine a target fitting circle; The target circle center coordinates are determined according to the target fitting circle.

10. A spherical or quasi-spherical object positioning device based on point cloud information, characterized in that: include: an acquisition unit, configured to acquire an image and a depth map of a spherical or quasi-spherical object, wherein the image of the spherical or quasi-spherical object includes a plurality of pixels, and the depth map of the spherical or quasi-spherical object is used to indicate a depth value of each of the pixels; A segmentation unit, used for performing semantic segmentation on the image of the spherical or quasi-spherical object to obtain a target pixel area; a point cloud determination unit, configured to map the target pixel area to a depth map of the spherical or quasi-spherical object to determine a target point cloud; The point set determination unit is used to perform equal depth slicing processing on the target point cloud based on a preset target depth value to obtain a point set to be fitted; wherein the steps for determining the preset target depth value are as follows: According to different types of the spherical or quasi-spherical objects, setting a depth selection ratio of equal-depth slices of the spherical or quasi-spherical objects within a preset threshold range; Determining a depth range of the target point cloud; The depth slice depth value is calculated by the depth selection ratio and the depth range; The positioning unit is used to perform circle detection on the set of points to be fitted, and determine the target fitting circle and the target circle center coordinates corresponding to the target fitting circle.

11. 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 method for locating a spherical or quasi-spherical object based on point cloud information as described in any one of claims 1 to 9 when executing the computer program.

12. A computer-readable storage medium storing computer-executable instructions, characterized in that: The computer executable instructions are used to execute the method for locating a spherical or quasi-spherical object based on point cloud information as described in any one of claims 1 to 9.

Images