CN111935642B - Positioning method and device of movable equipment - Google Patents
Positioning method and device of movable equipment Download PDFInfo
- Publication number
- CN111935642B CN111935642B CN201910439536.6A CN201910439536A CN111935642B CN 111935642 B CN111935642 B CN 111935642B CN 201910439536 A CN201910439536 A CN 201910439536A CN 111935642 B CN111935642 B CN 111935642B
- Authority
- CN
- China
- Prior art keywords
- determining
- error parameter
- distance error
- unstructured
- structured
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/024—Guidance services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
Abstract
The invention discloses a positioning method of movable equipment, which comprises the following steps: determining objects in the point cloud data; carrying out parameter structurization on the object in the point cloud data to obtain a structured object and an unstructured object; determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data; determining a first distance error parameter of the structured object from the first object and determining a second distance error parameter of the unstructured object from the second object; a loss function is constructed based on the first and second range error parameters, and a position of the movable device is determined based on the loss function. The invention realizes the technical effects of improving the matching precision and reducing the calculated amount. Meanwhile, the invention also discloses a positioning device of the movable equipment, the electronic equipment and a computer readable storage medium.
Description
Technical Field
The invention relates to the technical field of navigation, in particular to a positioning method and device of movable equipment.
Background
Nowadays, navigation technology is widely popularized, and in a map application frequently used by people, a navigation function is integrated, which can guide a user to a place where the user wants to go.
In the navigation process, the vehicle needs to be matched and positioned in real time so as to determine the position of the vehicle in the map. However, the current matching positioning method has too low matching precision or too large calculation amount, and is difficult to meet the actual requirement.
Disclosure of Invention
The embodiment of the application provides a positioning method and a positioning device for a mobile device, solves the technical problems of low matching precision or large calculated amount in a matching positioning method in the prior art, and achieves the technical effects of improving the matching precision and reducing the calculated amount.
In a first aspect, the present application provides the following technical solutions through an embodiment of the present application:
a method of positioning a mobile device, comprising:
determining an object in the point cloud data;
carrying out parameter structurization on the object in the point cloud data to obtain a structured object and an unstructured object;
determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data;
determining a first distance error parameter of the structured object from the first object and determining a second distance error parameter of the unstructured object from the second object;
a loss function is constructed based on the first and second range error parameters, and a position of the movable device is determined based on the loss function.
In a second aspect, the present application provides the following technical solutions through an embodiment of the present application:
a positioning apparatus of a movable device, comprising:
a first determination unit for determining an object in the point cloud data;
the structuring unit is used for carrying out parameter structuring on the object in the point cloud data to obtain a structured object and an unstructured object;
the second determining unit is used for determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data;
a third determination unit for determining a first distance error parameter of the structured object from the first object and for determining a second distance error parameter of the unstructured object from the second object;
a construction unit for constructing a loss function based on the first and second distance error parameters and determining a position of the movable device based on the loss function.
In a third aspect, based on the same inventive concept, the present application provides the following technical solutions through an embodiment of the present application:
an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the method steps of any of the embodiments of the first aspect.
In a fourth aspect, based on the same inventive concept, the present application provides the following technical solutions through an embodiment of the present application:
a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, is adapted to carry out the method steps of any of the embodiments of the first aspect.
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
in the embodiment of the application, when matching and positioning are performed, semantic information is newly added as a matching element, a first object matched with a structured object and a second object matched with an unstructured object are determined in map data based on the semantic information, a first distance error parameter of the structured object and the first object and a second distance error parameter of the unstructured object and the second object are further determined, and a loss function is constructed to determine the position of the movable equipment, so that the matching and positioning accuracy is higher while the calculated amount is reduced, the technical problems of low matching accuracy or large calculated amount in a matching and positioning method in the prior art are solved, and the technical effects of improving the matching accuracy and reducing the calculated amount are achieved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
FIG. 1 is a flow chart of a method for positioning a mobile device according to an embodiment of the present disclosure;
FIG. 2 is a detailed flowchart of step S103 in the embodiment of the present application;
FIG. 3 is a block diagram of a positioning apparatus of a mobile device according to an embodiment of the present disclosure;
fig. 4 is a structural diagram of a second determination subunit 303 in the embodiment of the present application;
fig. 5 is a structural diagram of an electronic device in an embodiment of the present application.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
Summary of the application
As described above, in the prior art, when performing matching positioning, either the matching accuracy is too low or the calculation amount is too large, which makes it difficult to meet the actual requirements. The reason is as follows: if matching is simply performed based on point cloud (i.e. pixel points in the field image are matched with pixel points in the map), the calculation amount is too large. If the matching is performed based on objects (i.e. matching is performed by using objects in the live image and objects in the map), not all objects can be parameterized, so that the matching accuracy is not high.
Based on the technical problem, the basic idea of the application is to classify point cloud data based on semantic information in the point cloud data, determine a first object matched with a structured object and a second object matched with an unstructured object in map data, and then adopt different processing modes according to the characteristics of each object.
Specifically, the map construction method, the map construction device, the map construction apparatus, and the readable storage medium provided by the application determine an object in the point cloud data, perform parameter structuring on the object in the point cloud data to obtain a structured object and an unstructured object, determine a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the object in the point cloud data, determine a first distance error parameter between the structured object and the first object, determine a second distance error parameter between the unstructured object and the second object, and construct a loss function based on the first distance error parameter and the second distance error parameter, thereby determining the position of the mobile device.
When matching and positioning are carried out, the semantic information is added as a matching element, a first object matched with the structured object and a second object matched with the unstructured object are determined in the map data based on the semantic information, a first distance error parameter of the structured object and the first object and a second distance error parameter of the unstructured object and the second object are further determined, and a loss function is constructed to determine the position of the movable equipment, so that the matching and positioning accuracy is higher and the calculation amount is reduced.
Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.
Exemplary method
The embodiment provides a positioning method of a mobile device, which is applied to the mobile device, and the mobile device may be: the device comprises an automobile, an electric vehicle, a robot and the like, which can run under the drive of a driving device (such as an electric motor or an oil engine). Here, the embodiment is not particularly limited as to what kind of the removable device is.
As shown in fig. 1, the method for positioning a mobile device includes the following steps:
step S101: an object in the point cloud data is determined.
In the specific implementation process, a perception sensor (such as a laser radar, a millimeter wave radar or a camera) is arranged on the movable equipment, and in the driving process of the movable equipment, the perception sensor can scan objects on the current road in real time to obtain corresponding point cloud data.
The Point Cloud (Point Cloud) refers to a collection of a huge amount of points (i.e., sampling points) on the surface of a target object (i.e., a scanned object), wherein information such as the position, distance and the like of each Point is carried.
Specifically, the point cloud obtained according to the laser measurement principle includes three-dimensional coordinates (XYZ) and laser reflection Intensity (Intensity); a point cloud obtained according to photogrammetry principles, comprising three-dimensional coordinates (XYZ) and color information (RGB); and combining laser measurement and photogrammetry principles to obtain a point cloud comprising three-dimensional coordinates (XYZ), laser reflection Intensity (Intensity) and color information (RGB). Of course, the point cloud can also be constructed by only using a camera and adopting a machine vision method.
Furthermore, semantic segmentation is carried out on the obtained point cloud, and a target object in the point cloud and semantic category information corresponding to the target object are obtained. Here, the task of semantic segmentation is to segment an image and distinguish different segmentations. When semantic segmentation is used, the point cloud is divided into semantically meaningful parts, and then the individual constituent points of each part are labeled as one of the predetermined categories, thereby identifying the different objects within the point cloud data.
Since different objects have different geometric shapes, semantic information of an object corresponding to the geometric shape in the point cloud data can be obtained based on the geometric shapes.
For example, the utility pole is mostly cylindrical, the high building is mostly cuboid, the guideboard is mostly rectangular flat, and the like. After identifying the geometric shapes from the point cloud data, semantic information of objects in the point cloud data corresponding to the geometric shapes may be obtained based on the geometric shapes. For example, if an object is cylindrical, which may be a utility pole, the object may be semantically labeled "post"; if a certain object is a cuboid which may be a high-rise building, a semantic label 'building' can be added to the object; if an object is rectangular and flat, which may be a guideboard, the object may be tagged with the semantic label "traffic sign".
For example, a tree also has a basic shape (i.e., a tree shape), and if an object is recognized as a tree, the object may be labeled as a "tree" semantically.
Of course, the above are only a few simple examples, and in practical applications, the geometric shape of the object is more complicated.
Step S102: and carrying out parameter structuring on the object in the point cloud data to obtain a structured object and an unstructured object.
Before executing step S102, the point cloud data may be preprocessed, specifically including: and denoising the point cloud data. Noise inevitably exists in the point cloud data, the noise can affect the whole matching and positioning process, and the point cloud data can be obtained through denoising treatment.
In the implementation process, the objects included in the point cloud data can be classified into two types, one type is an object that can be structured by parameters (i.e., structured object), and the other type is an object that cannot be structured by parameters (i.e., unstructured object). The structured objects often have regular shapes (such as tall buildings, telegraph poles, bridges, roads on roads, traffic signs and the like), and parametric characteristic points of the objects can be extracted based on the existing algorithm, and corresponding structural equation expressions are obtained; and the unstructured objects are irregular in shape (for example, trees are complex in shape, and the shape of each tree is different), parametric characteristic points of the objects are difficult to extract, and corresponding structured equation expressions are difficult to obtain.
In a specific implementation process, objects included in the point cloud data may be classified based on semantic information (i.e., semantic tags) of each object in the point cloud data, so as to obtain structured objects and unstructured objects.
For example, an object whose semantic label is "building", "pillar", "lane line", "traffic sign" may be determined as a structured object, and an object whose semantic label is "tree" may be determined as an unstructured object.
Step S103: and determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data.
In a specific implementation process, the map data may be downloaded to the mobile device in advance, or may be loaded in real time on line, where the map data includes image information of the current location of the mobile device. In order to save the amount of computation, only the local map information related to the current position in the map data may be described.
In a specific implementation process, a one-to-one correspondence relationship between the objects in the point cloud data and the objects in the map data can be established based on the semantic information of the objects in the point cloud data, so that the objects corresponding to the point cloud data are found in the map data, and a plurality of pairs of objects with the correspondence relationship are obtained, wherein each pair of objects with the correspondence relationship has the same semantic information. An object in the map data that matches a structured object in the point cloud data is referred to as a first object, and an object in the map data that matches an unstructured object in the point cloud data is referred to as a second object. Here, it is necessary to find an object matching each object in the point cloud data in the map data for semantic information of the object. Because the first object and the second object are matched in the map data based on the semantic information of the objects in the point cloud data, the calculation amount can be saved, the matching rate is improved, and the matching precision is improved.
For example, the object S is included in the point cloud data 1 An object S 2 An object S 3 An object S 4 And an object S 5 Wherein the object S 1 An object S 2 An object S 3 An object S 4 For structuring objects, objects S 1 Semantic tag of (2) is "building", object S 2 The semantic label of (1) is a "pillar", an object S 3 The semantic label of (1) is 'lane line', object S 4 The semantic label of (1) is a "traffic sign",object S 5 Being an unstructured object, object S 5 The semantic label of (a) is a "tree".
Correspondingly, it is necessary to find in the map data: with the object S 1 Matched objects S 1 ' (object S) 1 And an object S 1 ' is a pair of objects having a corresponding relationship, and the semantic labels of the objects are ' buildings '), and the object S 2 Matched objects S 2 ' (object S) 2 And an object S 2 ' is a pair of objects having a corresponding relationship, and the semantic labels of the objects are both ' pillars '), and the object S 3 Matched objects S 3 ' (object S) 3 And an object S 3 ' is a pair of objects having a corresponding relationship, and the semantic labels of the objects are both ' lane lines '), and the object S 4 Matched objects S 4 ' (object S) 4 And an object S 4 ' is a pair of objects with corresponding relationship, the semantic labels of the two are ' traffic board '), and the object S 5 Matched objects S 5 ' (object S) 5 And an object S 5 ' is a pair of objects with corresponding relationship, and the semantic labels of the two objects are ' trees '). Therefore, the calculation amount can be saved, the matching rate can be improved, and the matching precision can be improved.
Step S104: a first distance error parameter of the structured object from the first object is determined, and a second distance error parameter of the unstructured object from the second object is determined.
In a specific implementation process, for each pair of objects (including a structural object and an unstructured object) having a corresponding relationship, a distance error parameter corresponding to each pair of objects may be obtained.
In a specific implementation process, for each structured object, extraction of a structured parameter may be performed, and then, based on the extracted structured parameter, an euclidean distance between the structured object and the first object is calculated, so as to obtain a first distance error parameter. The matching positioning mode based on the object is adopted, and the method has the advantage of small calculation amount.
For example, for a lane line, the corresponding structural equation expression Ax can be obtained 3 +Bx 2 +Cx+D＝And 0, wherein { A, B, C and D } is an equation parameter, and a distance error parameter corresponding to the lane line is calculated based on the structural equation expression.
In a specific implementation, for each unstructured object, the second distance error parameter may be obtained based on euclidean distances between a plurality of points in the unstructured object and a plurality of points corresponding to the second object. The matching positioning mode based on the point cloud is adopted, although the calculation amount is large, the matching precision is high.
Because the mode of matching and positioning based on the object and the mode of matching and positioning based on the point cloud are combined during matching and positioning, the advantages of the two matching modes are taken into account, the calculated amount is reduced, and the matching precision is improved. Therefore, the technical problems of low matching precision or large calculated amount in the matching positioning method in the prior art are solved, and the technical effects of improving the matching precision and reducing the calculated amount are achieved.
Step S105: a loss function is constructed based on the first and second range error parameters, and a position of the movable device is determined based on the loss function.
In a specific implementation, after obtaining distance error parameters corresponding to each pair of objects (including the structural object and the unstructured object) corresponding to each pair of objects, a loss function may be constructed based on the distance error parameters.
Each range error parameter herein (e.g., a first range error parameter and a second range error parameter) is used to form an error term of the loss function.
By way of example, the object S can be calculated 1 With the object S 1 ' distance error parameter R between 1 Calculating the object S 2 With the object S 2 ' distance error parameter R between 2 Calculating the object S 3 With the object S 3 ' distance error parameter R between 3 Calculating the object S 4 With the object S 4 ' distance error parameter R between 4 Calculating the object S 5 With the object S 5 ' betweenDistance error parameter R of 5 . Based on the distance error parameter R 1 、R 2 、R 3 、R 4 、R 5 A loss function is constructed.
In the specific implementation process, a Euclidean distance between a structured object in the point cloud data and a first object in the map data is calculated to obtain a first distance error parameter which is used as an error term of a loss function, and similarly, a Euclidean distance between an unstructured object in the point cloud data and a second object in the map data is calculated to obtain a second distance error parameter which is used as an error term of the loss function, so that a loss function is constructed, the minimized error of the loss function is solved, the positioning problem is converted into a solving problem of the minimized error term, the point cloud data and the map data are registered as much as possible, and the positioning result is obtained.
For example, for the distance error parameter R 1 、R 2 、R 3 、R 4 、R 5 A loss function may be constructed by obtaining a corresponding error term based on each distance error parameter and summing the obtained error terms in sequence. And then, the sum of all error terms is minimized through a plurality of iterations, so that the aim of minimizing the error terms is fulfilled, a registration result is obtained, and the position of the movable equipment can be determined.
In the embodiment of the application, when matching and positioning are performed, semantic information is newly added as a matching element, a first object matched with a structured object and a second object matched with an unstructured object are determined in map data based on the semantic information, a first distance error parameter of the structured object and the first object and a second distance error parameter of the unstructured object and the second object are further determined, and a loss function is constructed to determine the position of the movable equipment, so that the matching and positioning accuracy is higher while the calculated amount is reduced, the technical problems of low matching accuracy or large calculated amount in a matching and positioning method in the prior art are solved, and the technical effects of improving the matching accuracy and reducing the calculated amount are achieved.
As shown in fig. 2, on the basis of the embodiment shown in fig. 1, as an optional implementation manner of this embodiment, the step S103 may include the following steps:
step S201: determining a first object which has the same semantic information with the structured object and is closest to the structured object in the map data; and
step S202: and determining a second object which has the same semantic information with the unstructured object and is closest to the unstructured object in the map data.
In a specific implementation process, the position of each object can be represented by the central point of each object, and when the closest object is associated, the closest central point of the same kind of object is searched for association. And selecting a pair of objects with the same semantic information and the closest distance from the point cloud data and the map data to be associated by taking the semantic information as a new dimension. Wherein the distance is in particular a euclidean distance (i.e. a euclidean distance).
For example, for a structured object S in point cloud data 1 If the semantic information is 'building', the map data is searched for the semantic information which is 'building' and is similar to the structural object S 1 And the object closest to the euclidean distance of (c) is taken as the structured object S 1 First matched objects S 1 ’。
For example, for unstructured objects S in point cloud data 4 If the semantic information is tree, searching the map data for the semantic information which is tree and is similar to the structured object S 4 And the object with the closest Euclidean distance is taken as the unstructured object S 4 First matched objects S 4 ’。
In this way, after steps S201 and S202, the objects in the point cloud data and the objects in the map data can be associated one by one.
In one example, the determining a first distance error parameter of the structured object from the first object includes:
the first method is as follows: determining a first reference line on the structured object and a second reference line on the first object matching the first reference line; calculating the Euclidean distance between the first reference line and the second reference line to obtain a first distance error parameter; or
The second method comprises the following steps: determining a first reference surface from the structured object, and determining a second reference surface matched with the first reference surface from the first object; and calculating the Euclidean distance between the first reference surface and the second reference surface to obtain a first distance error parameter.
The first and second selection modes depend on the geometric shape of the matched object. For example, the signboard and the traffic light can determine the reference surface, i.e. the second mode is adopted; the lane line and the pillar can determine the reference line, namely, the first mode is adopted.
In a specific implementation, with reference to the above manner one, a "line-to-line" distance error parameter may be calculated for the structured object. Specifically, a reference line (i.e., a first reference line) may be selected on each structured object in the point cloud data, another reference line (i.e., a second reference line) matching the reference line may be found in the map data, and the euclidean distance between the first reference line and the second reference line may be calculated to obtain the distance error parameters (i.e., first distance error parameters) corresponding to the pair of objects.
In a specific implementation process, with reference to the second method, a "plane-to-plane" distance error parameter may be calculated for the structured object. Specifically, a reference surface (i.e., a first reference surface) may be selected on each structured object in the point cloud data, another reference surface (i.e., a second reference surface) matching the reference surface may be found in the map data, and the euclidean distance between the first reference surface and the second reference surface may be calculated to obtain a distance error parameter (i.e., a first distance error parameter) corresponding to the pair of objects. In this embodiment, for a structured object, when determining its corresponding distance error parameter, it may be based on a "line-to-line" or "surface-to-surface" approach, thereby reducing the amount of computation.
In one example, the determining a second distance error parameter of the unstructured object from the second object comprises:
determining M first reference points on the unstructured object and M second reference points on the second object, wherein the M first reference points are matched with the M second reference points one by one to form M reference point pairs, and M is an integer greater than or equal to 2; determining a second distance error parameter based on Euclidean distances between each of the M pairs of reference points.
In a specific implementation process, M first reference points may be determined on an unstructured object, then points on a second object are retrieved within a search radius, and the only determined nearest M second reference points are found, where the M second reference points are paired with the M first reference points one by one to form M reference point pairs.
In particular embodiments, a preferred value for M is 6. 6 is preferred here because it is considered that if there are no structured objects present, only unstructured objects, then at least 6 point pairs are needed to solve for the current pose of the mobile device.
In a specific implementation process, for point cloud data, an error term of the point cloud data is formed by Euclidean distances between matched nearest neighbor point pairs. In order to avoid the imbalance of error items caused by the inconsistent number of point clouds, the single error item provided by the scheme is for an object rather than for a single point. That is, the sum of euclidean distances between matching nearest neighbor point pairs is calculated for the point cloud data, and then divided by the total number of the point pairs to obtain a value as an error term of the object.
For example, for an unstructured object S 4 And a second object S 4 ', can be on the unstructured object S 4 6 reference points are selected, and are respectively P 1 、P 2 、P 3 、P 4 、P 5 、P 6 Searching for and P on a second object 1 The nearest reference point is P 1 ’(P 1 And P 1 ' match, form a pair of points), and similarly, search for P on the second object 2 The nearest reference point is P 2 ’(P 2 And P 2 ' match, form a pair of points) on a second object to obtain a sum P 3 The nearest reference point is P 3 ’(P 3 And P 3 Are matched to form aPoint pair), searching for and obtaining P on a second object 4 The nearest reference point is P 4 ’(P 4 And P4' are matched to form a point pair), searching and obtaining the point pair with P on the second object 5 The nearest reference point is P 5 ’(P 5 And P 5 ' match, form a pair of points) on a second object to obtain a sum P 6 The nearest reference point is P 6 ’(P 6 And P 6 ' match, constitute a pair of points).
Further, P is calculated 1 And P 1 ' Euclidean distance L 1 Calculate P 2 And P 2 ' Euclidean distance L 2 Calculate P 3 And P 3 ' Euclidean distance L 3 Calculate P 4 And P 4 ' Euclidean distance L 4 Calculating P 5 And P 5 ' Euclidean distance L 5 Calculate P 6 And P 6 ' Euclidean distance L 6 . Finally, mixing L 1 L6 are sequentially added and then divided by 6, namely the unstructured object S is obtained 4 And a corresponding second distance error parameter, which is an error term of the loss function.
In one example, said constructing a loss function based on said first distance error parameter and said second distance error parameter comprises:
distributing a weighted value for the first distance error parameter based on semantic information of the structured object; distributing a weighted value for the second distance error parameter based on the semantic information of the unstructured object; and constructing a loss function based on the first distance error parameter and the weight value thereof, and the second distance error parameter and the weight value thereof.
In a specific implementation, the weight γ may be configured for the first distance error parameter 1 Configuring a weight gamma for the second range error parameter 2 The constructed loss function can better accord with the actual condition, thereby being beneficial to obtaining more accurate and more reasonable loss functions. Wherein the weight may be determined by the referential of the corresponding object itself. For example, in the actual intersection, the artificial signs such as lane lines and signboard are more involvedFor consideration, the corresponding weight is also higher. And the tree and the structured high-rise have lower referential property and lower corresponding weight.
The technical scheme in the embodiment of the application at least has the following technical effects or advantages:
in the embodiment of the application, when matching and positioning are performed, semantic information is newly added as a matching element, a first object matched with a structured object and a second object matched with an unstructured object are determined in map data based on the semantic information, a first distance error parameter of the structured object and the first object and a second distance error parameter of the unstructured object and the second object are further determined, and a loss function is constructed to determine the position of the movable equipment, so that the matching and positioning accuracy is higher while the calculated amount is reduced, the technical problems of low matching accuracy or large calculated amount in a matching and positioning method in the prior art are solved, and the technical effects of improving the matching accuracy and reducing the calculated amount are achieved.
Exemplary devices
Based on the same inventive concept, as shown in fig. 3, the present embodiment provides a positioning apparatus 300 for a mobile device, comprising:
a first determination unit 301 configured to determine an object in the point cloud data;
a structuring unit 302, configured to perform parameter structuring on the object in the point cloud data, so as to obtain a structured object and an unstructured object;
a second determining unit 303, configured to determine, based on semantic information corresponding to an object in the point cloud data, a first object that matches the structured object and a second object that matches the unstructured object in the map data;
a third determining unit 304 for determining a first distance error parameter of the structured object from the first object and for determining a second distance error parameter of the unstructured object from the second object;
a construction unit 305 for constructing a loss function based on the first distance error parameter and the second distance error parameter, and determining the position of the movable device based on the loss function.
In one example, as shown in fig. 4, the second determining unit 303 includes:
a first determining subunit 3031, configured to determine, in the map data, a first object that has the same semantic information as the structured object and is closest to the structured object; and
a second determining subunit 3032, configured to determine, in the map data, a second object that has the same semantic information as the unstructured object and is closest to the unstructured object.
In one example, the third determining unit 304 includes:
a third determining subunit for determining a first reference line on the structured object and a second reference line on the first object matching the first reference line; calculating the Euclidean distance between the first reference line and the second reference line to obtain a first distance error parameter; or
A fourth determining subunit, configured to determine a first reference surface from the structured object, and determine a second reference surface matching the first reference surface from the first object; and calculating the Euclidean distance between the first reference surface and the second reference surface to obtain the first distance error parameter.
In an example, the third determining unit 304 further includes:
a fifth determining subunit, configured to determine M first reference points on the unstructured object and M second reference points on the second object, where the M first reference points and the M second reference points are matched one to form M reference point pairs, and M is an integer greater than or equal to 2; determining the second distance error parameter based on Euclidean distances between each of the M reference point pairs.
In one example, the building unit 305 includes:
a first assigning subunit, configured to assign a weight value to the first distance error parameter based on semantic information of the structured object;
a second assigning subunit, configured to assign a weight value to the second distance error parameter based on semantic information of the unstructured object;
a construction subunit, configured to construct the loss function based on the first distance error parameter and the weight value thereof, and the second distance error parameter and the weight value thereof.
It will be understood by those skilled in the art herein that the detailed functions and operations of the respective units/modules in the above-described positioning apparatus 300 for a mobile device have been described in detail in the above description of the positioning method for a mobile device with reference to fig. 1 to 2, and thus, a repetitive description thereof will be omitted.
As described above, the positioning apparatus 300 of a movable device according to an embodiment of the present application may be implemented in a device terminal of various movable devices, for example, a computer or a microprocessor used in an autonomous vehicle, a robot, or the like. In one example, the positioning apparatus 300 of the removable device according to the embodiment of the present application may be integrated into a device terminal of the removable device as a software module and/or a hardware module. For example, the positioning apparatus 300 of the mobile device may be a software module in an operating system of the device terminal, or may be an application developed for the device terminal; of course, the positioning means 300 of the removable device may equally be one of many hardware modules of the terminal of the device.
Alternatively, in another example, the positioning apparatus 300 of the mobile device and the device terminal of the mobile device may also be separate devices, and the positioning apparatus 300 of the mobile device may be connected to the device terminal through a wired and/or wireless network and transmit the interactive information according to an agreed data format.
Exemplary electronic device
Based on the same inventive concept, an electronic apparatus according to an embodiment of the present application is described below with reference to fig. 5.
FIG. 5 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.
As shown in fig. 5, the electronic device 500 includes one or more processors 501 and memory 502.
The processor 501 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 500 to perform desired functions.
Memory 502 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium and executed by the processor 501 to implement the positioning method of the removable device of the various embodiments of the present application described above and/or other desired functions. Various contents such as an input signal, a signal component, a noise component, etc. may also be stored in the computer-readable storage medium.
In one example, the electronic device 500 may further include: an input device 503 and an output device 504, which are interconnected by a bus system and/or other form of connection mechanism (not shown).
The input device 503 may also include, for example, a keyboard, a mouse, and the like.
The output device 504 may output various information to the outside, including the determined distance information, direction information, and the like. The output devices 504 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.
Of course, for simplicity, only some of the components of the electronic device 500 relevant to the present application are shown in fig. 5, and components such as buses, input/output interfaces, and the like are omitted. In addition, electronic device 500 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer-readable storage Medium
Based on the same inventive concept, in addition to the above described methods and devices, embodiments of the present application may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform the steps in the method for positioning a movable device according to various embodiments of the present application described in the above described "exemplary methods" section of this specification.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform steps in a method of positioning a movable apparatus according to various embodiments of the present application described in the "exemplary methods" section above in this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
Claims (10)
1. A method of positioning a mobile device, comprising:
determining an object in the point cloud data;
carrying out parameter structuralization on the objects in the point cloud data to obtain structured objects and unstructured objects;
determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data;
determining a first distance error parameter of the structured object from the first object and a second distance error parameter of the unstructured object from the second object;
a loss function is constructed based on the first and second range error parameters, and a position of the movable device is determined based on the loss function.
2. The method of claim 1, the determining a first object in the map data that matches the structured object and a second object that matches the unstructured object based on semantic information corresponding to objects in the point cloud data, comprising:
determining a first object which has the same semantic information as the structured object and is closest to the structured object in the map data; and
and determining a second object which has the same semantic information with the unstructured object and is closest to the unstructured object in the map data.
3. The method of claim 2, the determining a first distance error parameter of the structured object from the first object, comprising:
determining a first reference line on the structured object and a second reference line on the first object matching the first reference line; calculating the Euclidean distance between the first reference line and the second reference line to obtain a first distance error parameter; or
Determining a first reference surface from the structured object and a second reference surface from the first object matching the first reference surface; and calculating the Euclidean distance between the first reference surface and the second reference surface to obtain the first distance error parameter.
4. The method of claim 2, the determining a second distance error parameter for the unstructured object and the second object, comprising:
determining M first reference points on the unstructured object and M second reference points on the second object, wherein the M first reference points are matched with the M second reference points one by one to form M reference point pairs, and M is an integer greater than or equal to 2;
determining the second distance error parameter based on Euclidean distances between each of the M reference point pairs.
5. The method of any of claims 1-4, wherein constructing a loss function based on the first distance error parameter and the second distance error parameter comprises:
based on semantic information of the structured object, assigning a weight value to the first distance error parameter;
assigning a weight value to the second distance error parameter based on semantic information of the unstructured object;
and constructing the loss function based on the first distance error parameter and the weight value thereof, and the second distance error parameter and the weight value thereof.
6. A positioning apparatus of a movable device, comprising:
a first determination unit for determining an object in the point cloud data;
the structuring unit is used for carrying out parameter structuring on the object in the point cloud data to obtain a structured object and an unstructured object;
the second determining unit is used for determining a first object matched with the structured object and a second object matched with the unstructured object in the map data based on semantic information corresponding to the objects in the point cloud data;
a third determination unit for determining a first distance error parameter of the structured object from the first object and for determining a second distance error parameter of the unstructured object from the second object;
a construction unit configured to construct a loss function based on the first distance error parameter and the second distance error parameter, and determine a position of the movable device based on the loss function.
7. The apparatus of claim 6, the second determination unit, comprising:
a first determining subunit, configured to determine, in the map data, a first object that has the same semantic information as the structured object and is closest to the structured object; and
and the second determining subunit is used for determining a second object which has the same semantic information with the unstructured object and is closest to the unstructured object in the map data.
8. The apparatus of any of claims 6 to 7, the building unit comprising:
a first assigning subunit, configured to assign a weight value to the first distance error parameter based on semantic information of the structured object;
a second assigning subunit, configured to assign a weight value to the second distance error parameter based on semantic information of the unstructured object;
a construction subunit, configured to construct the loss function based on the first distance error parameter and the weight value thereof, and the second distance error parameter and the weight value thereof.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the method steps as claimed in any of claims 1 to 5.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, is adapted to carry out the method steps of any of claims 1 to 5.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2019103953990 | 2019-05-13 | ||
| CN201910395399 | 2019-05-13 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111935642A CN111935642A (en) | 2020-11-13 |
| CN111935642B true CN111935642B (en) | 2022-10-21 |
Family
ID=73282929
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910439536.6A Active CN111935642B (en) | 2019-05-13 | 2019-05-24 | Positioning method and device of movable equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111935642B (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016118672A2 (en) * | 2015-01-20 | 2016-07-28 | Solfice Research, Inc. | Real time machine vision and point-cloud analysis for remote sensing and vehicle control |
| US10684372B2 (en) * | 2017-10-03 | 2020-06-16 | Uatc, Llc | Systems, devices, and methods for autonomous vehicle localization |
| CN108198145B (en) * | 2017-12-29 | 2020-08-28 | 百度在线网络技术（北京）有限公司 | Method and device for point cloud data restoration |
| CN109345574B (en) * | 2018-08-31 | 2020-10-09 | 西安电子科技大学 | Laser radar three-dimensional mapping method based on semantic point cloud registration |
| CN109737974B (en) * | 2018-12-14 | 2020-11-27 | 中国科学院深圳先进技术研究院 | 3D navigation semantic map updating method, device and equipment |
-
2019
- 2019-05-24 CN CN201910439536.6A patent/CN111935642B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111935642A (en) | 2020-11-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6862409B2 (en) | Map generation and moving subject positioning methods and devices | |
| EP3961579A2 (en) | Target detection method, apparatus, medium and computer program product | |
| CN111694903B (en) | Map construction method, device, equipment and readable storage medium | |
| CN110533055B (en) | Point cloud data processing method and device | |
| CN109872392B (en) | Man-machine interaction method and device based on high-precision map | |
| EP4177836A1 (en) | Target detection method and apparatus, and computer-readable medium and electronic device | |
| CN113859264B (en) | Vehicle control method, device, electronic equipment and storage medium | |
| CN111881777B (en) | Video processing method and device | |
| CN115855084A (en) | Map data fusion method and device, electronic equipment and automatic driving product | |
| CN113724388B (en) | High-precision map generation method, device, equipment and storage medium | |
| CN114186007A (en) | High-precision map generation method and device, electronic equipment and storage medium | |
| US11699234B2 (en) | Semantic segmentation ground truth correction with spatial transformer networks | |
| CN111935642B (en) | Positioning method and device of movable equipment | |
| CN110363847B (en) | Map model construction method and device based on point cloud data | |
| CN116844129A (en) | Road side target detection method, system and device for multi-mode feature alignment fusion | |
| CN115062240A (en) | Parking lot sorting method and device, electronic equipment and storage medium | |
| CN114140813A (en) | High-precision map marking method, device, equipment and storage medium | |
| CN110647599B (en) | Method and apparatus for displaying information | |
| CN114419564A (en) | Vehicle pose detection method, device, equipment, medium and automatic driving vehicle | |
| CN114187357A (en) | High-precision map production method and device, electronic equipment and storage medium | |
| CN110120075B (en) | Method and apparatus for processing information | |
| CN112987707A (en) | Automatic driving control method and device for vehicle | |
| CN113762397A (en) | Detection model training and high-precision map updating method, device, medium and product | |
| CN113654548A (en) | Positioning method, positioning device, electronic equipment and storage medium | |
| CN116168366B (en) | Point cloud data generation method, model training method, target detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |