A large-range trajectory potential region estimation method and device considering terrain features

Abstract

The application discloses a large-range trajectory potential area estimation method and device considering terrain features, the method obtains terrain feature information corresponding to an area where a target user trajectory is located, and obtains trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm; the terrain feature information and the trajectory feature information are input into a hybrid deep learning network to obtain uncertainty trajectory errors corresponding to each position point on the target user trajectory respectively; and a potential area corresponding to the target user trajectory is determined according to the terrain feature information and the uncertainty trajectory errors. Since the potential area determined according to the terrain feature information and the trajectory feature information considers the complexity of the environment and the error caused by the global positioning system update, the problem that the predicted motion trajectory potential area is not accurate enough due to the fact that the existing potential area prediction model ignores the time-varying measurement error of the global positioning system and the influence of the urban terrain environment is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of positioning, in particular to a large-range trajectory potential area estimation method and device considering terrain features. BACKGROUND

[0002] Location-based services play an indispensable role in the application of intelligent terminals, and how to accurately obtain continuous and reliable position information of users in a large range of scenarios becomes the key. Modeling the potential area of pedestrian motion trajectory in a complex urban environment is considered to be a meaningful and challenging task to promote geospatial data mining and analysis.

[0003] At present, due to the complexity and diversity of urban scenes, the existing positioning technology and potential area estimation technology based on intelligent terminals in a large range of urban scenes mostly only consider the motion distance or speed, ignoring the influence of time-varying measurement error of the global positioning system and urban terrain environment on potential area estimation, resulting in that the predicted motion trajectory potential area of the existing user motion trajectory potential area prediction model is not accurate enough.

[0004] Therefore, the prior art still needs to be improved and developed. SUMMARY

[0005] The technical problem to be solved by the present application is that, in view of the above defects of the prior art, a large-range outdoor trajectory potential area estimation method and device considering terrain features are provided, aiming to solve the problem that the predicted motion trajectory potential area of the existing user motion trajectory potential area prediction model is not accurate enough due to the neglect of the influence of time-varying measurement error of the global positioning system and urban terrain environment.

[0006] The technical scheme adopted by the present application to solve the problem is as follows:

[0007] In a first aspect, the present application provides a large-range trajectory potential area estimation method considering terrain features, wherein the method comprises:

[0008] Obtaining terrain feature information corresponding to the area where the target user trajectory is located, wherein the terrain feature information includes the slope, slope direction and slope height of each position point on the target user trajectory;

[0009] Obtaining trajectory feature information determined based on the global positioning system and the pedestrian dead reckoning algorithm, wherein the trajectory feature information includes information for describing the change of the target user trajectory determined based on the global positioning system and the pedestrian dead reckoning algorithm respectively;

[0010] inputting the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain an uncertainty trajectory error corresponding to each of the position points on the target user trajectory respectively;

[0011] determining a potential area corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory error.

[0012] In an implementation method, the trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm includes:

[0013] obtaining first trajectory feature information determined based on the global positioning system, wherein the first trajectory feature information includes a first straight-line distance difference, a first coordinate change amount, a first reference speed, a first virtual heading value, a signal-to-noise ratio average of satellite data, and a global positioning system update time determined based on the global positioning system;

[0014] obtaining second trajectory feature information determined based on the pedestrian dead reckoning algorithm, wherein the second trajectory feature information includes a step heading vector, a second straight-line distance difference, a second coordinate change amount, and a second reference speed determined based on the pedestrian dead reckoning algorithm;

[0015] determining third trajectory feature information according to the first trajectory feature information and the second trajectory feature information;

[0016] determining the trajectory feature information according to the first trajectory feature information, the second trajectory feature information, and the third trajectory feature information.

[0017] In an implementation method, the determination of the third trajectory feature information according to the first trajectory feature information and the second trajectory feature information includes:

[0018] determining a straight-line distance residual according to the first straight-line distance difference and the second straight-line distance difference;

[0019] taking the straight-line distance residual as the third trajectory feature information.

[0020] In an implementation method, the hybrid deep learning network includes a bidirectional long short-term memory network and a multilayer perceptron.

[0021] In an implementation method, the determination of the potential area corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory error includes:

[0022] obtaining a coordinate position corresponding to each of the position points on the target user trajectory;

[0023] construct a potential region corresponding to each of the position points according to the coordinate positions, the terrain feature information, and the uncertainty trajectory error corresponding to each of the position points;

[0024] calculate a union region of the potential regions corresponding to each of the position points, and take the union region as a potential region corresponding to the target user trajectory.

[0025] In one implementation method, the constructing of the potential region corresponding to each of the position points according to the coordinate positions, the terrain feature information, and the uncertainty trajectory error corresponding to each of the position points comprises:

[0026] when the slope in the terrain feature information corresponding to the position point is not equal to zero, determining that the potential region is an ellipse;

[0027] determining an ellipse according to the terrain feature information, taking the position coordinate of the position point as the center of the ellipse, and taking the uncertainty trajectory error corresponding to the position point as the minor axis value of the ellipse, to determine the potential region corresponding to the position point.

[0028] In one implementation method, the method further comprises:

[0029] when the slope in the terrain feature information corresponding to the position point is equal to zero, determining that the potential region is a circle;

[0030] taking the position coordinate of the position point as the center of the circle, and taking the uncertainty trajectory error corresponding to the position point as the radius of the circle, to determine the potential region corresponding to the position point.

[0031] In a second aspect, the embodiments of the present application further provide a large-range trajectory potential region estimation device considering terrain features, wherein the large-range trajectory potential region estimation device considering terrain features comprises:

[0032] a terrain feature acquisition module configured to acquire terrain feature information corresponding to a region where a target user trajectory is located, wherein the terrain feature information comprises the slope, slope direction, and slope height of each position point on the target user trajectory;

[0033] a trajectory feature acquisition module configured to acquire trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm, wherein the trajectory feature information comprises information determined based on the global positioning system and the pedestrian dead reckoning algorithm respectively to describe the change of the target user trajectory;

[0034] a trajectory error learning module configured to input the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain an uncertainty trajectory error corresponding to each of the position points on the target user trajectory respectively.

[0035] a potential region determining module configured to determine a potential region corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory errors.

[0036] In a third aspect, an embodiment of the present application further provides a terminal, wherein the terminal comprises a memory and one or more processors; the memory stores one or more programs; the programs contain instructions for executing the method for estimating a large-scale trajectory potential region considering terrain features according to any one of the above solutions; and the processors are configured to execute the programs.

[0037] In a fourth aspect, an embodiment of the present application further provides a computer readable storage medium, which stores a plurality of instructions, and the instructions are suitable for being loaded and executed by a processor to implement the steps of the method for estimating a large-scale trajectory potential region considering terrain features according to any one of the above solutions.

[0038] The present application has the following beneficial effects: the embodiment of the present application obtains terrain feature information corresponding to a region where a target user trajectory is located, and obtains trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm; inputs the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain uncertainty trajectory errors corresponding to each position point on the target user trajectory respectively; and determines a potential region corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory errors. Since the potential region determined according to the terrain feature information and the trajectory feature information takes into account the complexity of the environment and the error caused by the update of the global positioning system, the problem that the predicted motion trajectory potential region is not accurate enough due to the neglect of the time-varying measurement error of the global positioning system and the influence of the urban terrain environment in the existing potential region prediction model is solved. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0040] Figure 1 is a flow chart of the specific implementation of the method for estimating a large-scale trajectory potential region considering terrain features provided by the embodiment of the present application.

[0041] Figure 2 is a pixel representation diagram of the planar method provided by the embodiment of the present application.

[0042] Figure 3A mixed deep learning network structure diagram provided by an embodiment of the present application.

[0043] Figure 4 A principle diagram of a large-range trajectory potential region estimation device considering terrain features provided by an embodiment of the present application.

[0044] Figure 5 A principle diagram of a terminal provided by an embodiment of the present application. DETAILED DESCRIPTION

[0045] The present application discloses a large-range trajectory potential region estimation method and device considering terrain features, in order to make the purpose, technical scheme and effect of the present application more clear and explicit, the present application is further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0046] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "a", "an" and "the" used herein also include the plural forms. It should be further understood that the use of the term "include" in the specification of the present application means that the features, integers, steps, operations, elements and / or components exist, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we say that an element is "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or there can be intermediate elements. In addition, "connected" or "coupled" used herein can include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any single unit and all combinations of the associated listed items.

[0047] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as that generally understood by those skilled in the art to which the present application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have meanings consistent with those in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as such.

[0048] Location-based services play an indispensable role in the application of intelligent terminals, and how to accurately obtain continuous and reliable position information of users in a large range of scenarios becomes the key. Modeling the potential region of pedestrian motion trajectory in a complex urban environment is considered to be a meaningful and challenging task to promote geospatial data mining and analysis.

[0049] Currently, due to the complexity and diversity of urban scenes, the existing positioning technology and potential area estimation technology based on intelligent terminals in a wide range of urban scenes only consider the motion distance or speed, ignore the time-varying measurement error of the global positioning system and the influence of the urban terrain environment on the potential area estimation, resulting in that the predicted motion trajectory potential area of the existing user motion trajectory potential area prediction model is not accurate enough.

[0050] In view of the above defects of the prior art, the present application provides a wide-range trajectory potential area estimation method considering terrain features, which obtains terrain feature information corresponding to the area where the target user trajectory is located, and obtains trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm; inputs the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain the uncertainty trajectory error corresponding to each position point on the target user trajectory; and determines the potential area corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory error. Since the potential area determined according to the terrain feature information and the trajectory feature information in the present application considers the complexity of the environment and the error caused by the update of the global positioning system, it solves the problem that the predicted motion trajectory potential area is not accurate enough due to the neglect of the time-varying measurement error of the global positioning system and the influence of the urban terrain environment in the existing potential area prediction model.

[0051] Exemplary method

[0052] The present embodiment provides a wide-range trajectory potential area estimation method considering terrain features, which can be applied to a terminal device for predicting and estimating the potential area of a target user motion trajectory. The terminal device includes computer, smart phone, Internet of Things terminal and other intelligent product terminals. By collecting, processing and fusing multi-sensor and wireless signal data, the target user trajectory is reconstructed, and the potential area of the target user trajectory is estimated according to the target user trajectory.

[0053] As shown in Figure 1 The method comprises:

[0054] Step S100, obtaining terrain feature information corresponding to the area where the target user trajectory is located, wherein the terrain feature information includes the slope, slope direction and slope height of each position point on the target user trajectory.

[0055] Specifically, the target user trajectory in the embodiment represents the motion trajectory generated by the target user when moving or walking. Due to the complexity and diversity of the urban environment, the terrain feature information corresponding to each position point in the target user trajectory, that is, the slope, slope direction and slope height of the terrain, will affect the potential area size of the target user trajectory. Generally speaking, the slope is an index of the inclination degree of the local ground surface in space, and the size directly affects the scale and intensity of the ground material flow and energy conversion. The smaller the slope value is, the flatter the terrain is; the larger the slope value is, the steeper the terrain is, and the slope is an important factor for engineering design and productivity spatial layout. In addition, the slope has directionality, and the slope calculated based on the DEM is the slope value in the direction of the maximum slope. The embodiment obtains the terrain feature information corresponding to the area where the target user trajectory is located, and the terrain feature information includes the slope, slope direction and slope height of each position point on the target user trajectory, so as to estimate the potential area of the target user trajectory according to the terrain feature information, and ensure that the estimation result of the potential area of the target user trajectory is more accurate.

[0056] The following two methods are provided in the embodiment to realize slope calculation:

[0057] 1) Plane method

[0058] The slope calculation of the plane method is the surface change rate (increment) in the horizontal (dz / dx) and vertical (dz / dy) directions from the center pixel to each adjacent pixel. The values of the center pixel and its eight adjacent pixels determine the horizontal increment and the vertical increment. These adjacent pixels are identified using letters A to I, where E represents the pixel whose slope is currently being calculated, as shown in Figure 2 .

[0059] The length of the slope is:

[0060]

[0061] The angle of the slope is:

[0062]

[0063] 2) Geodesic method

[0064] The geodesic method measures the slope in the 3D geocentric coordinate system (also known as the Earth-Centered Earth-Fixed (ECEF) coordinate system) by regarding the shape of the earth as an ellipsoid. The calculation result is not affected by the projection method of the data set. It should be noted that if the unit of the height direction is defined in the spatial reference system, the unit of the height direction of the input grid is used in this method; if the unit of the height direction is not defined in the input spatial reference system, the unit of the height direction needs to be defined.

[0065] The geodesic calculation in this embodiment uses geodetic coordinates (latitude longitude λ, and height h) to calculate the X, Y, Z values of the ECEF coordinates. If the input surface raster is in a projected coordinate system (PCS), it needs to be re-projected to a geographic coordinate system (GCS) where each location has geodetic coordinates, which are then converted to the ECEF coordinate system. The height h (i.e., the Z value) is the ellipsoidal height with respect to the ellipsoid surface. The conversion from geodetic coordinates to ECEF coordinates is as follows:

[0066]

[0067]

[0068]

[0069] where, is the latitude, λ is the longitude, h is the ellipsoidal height, a is the ellipsoid major axis, and b is the ellipsoid minor axis.

[0070] The slope in the geodesic line refers to the angle formed between the terrain surface and the ellipsoid surface. For any surface parallel to the ellipsoid surface, the slope is 0. To calculate the slope at each location, a 3-pixel by 3-pixel neighborhood plane can be fitted around each processing pixel using the least squares method (LSM). In this embodiment, the least squares method (LSM) is used to minimize the sum of the squared differences between the actual height measurements and the fitted height measurements.

[0071] After fitting the plane, the surface normal is calculated at the pixel location. At the same location, the ellipsoidal normal that is perpendicular to the tangent plane of the ellipsoid surface is also calculated. The slope is calculated through the angle (β) between the ellipsoidal normal and the terrain surface normal and the percent of elevation gain:

[0072] Slope_PercentRise = tan -1 β * 100%

[0073] At the same time, the aspect information of the location points on the target user's trajectory is calculated, which is used to identify the compass direction that each location (DEM pixel) faces in the downhill slope. The aspect is represented by a positive number between 0 and 360 degrees, measured clockwise with north as the reference direction. Flat (with zero slope) pixels are assigned an aspect of -1.

[0074] As Figure 1 shown, the method further comprises the following steps:

[0075] In step S200, the trajectory feature information determined based on the global positioning system and the pedestrian dead reckoning algorithm is acquired, wherein the trajectory feature information includes information determined based on the global positioning system and the pedestrian dead reckoning algorithm respectively for describing the change of the trajectory of the target user.

[0076] Specifically, during the process of the target user completing the target user trajectory, there is a certain time-varying measurement error between the information (such as position, time, etc.) of the last position point measured by the global positioning system (GPS) and the information of the next position point. In this embodiment, the global positioning system is used to acquire the trajectory feature information of the target user trajectory, and the pedestrian dead reckoning algorithm is used to calculate the trajectory feature information of the target user trajectory. The change of the motion trajectory of the target user is described by the trajectory feature information determined based on the global positioning system and the pedestrian dead reckoning algorithm respectively, the richness of the data of the trajectory feature information is increased, the time measurement error caused by the data acquisition of the global positioning system is reduced, and the accuracy of the potential area estimation of the target user trajectory is improved.

[0077] In an implementation manner, the acquiring the trajectory feature information determined based on the global positioning system and the pedestrian dead reckoning algorithm comprises:

[0078] In step S201, the first trajectory feature information determined based on the global positioning system is acquired, wherein the first trajectory feature information includes a first straight line distance difference, a first coordinate change amount, a first reference speed, a first virtual heading value, a signal-to-noise ratio mean value of satellite data, and a global positioning system update time determined based on the global positioning system.

[0079] In step S202, the second trajectory feature information determined based on the pedestrian dead reckoning algorithm is acquired, wherein the second trajectory feature information includes a step heading vector, a second straight line distance difference, a second coordinate change amount, and a second reference speed determined based on the pedestrian dead reckoning algorithm.

[0080] In step S203, the third trajectory feature information is determined according to the first trajectory feature information and the second trajectory feature information.

[0081] In step S204, the trajectory feature information is determined according to the first trajectory feature information, the second trajectory feature information, and the third trajectory feature information.

[0082] Specifically, the embodiment obtains a first straight line distance difference, a first coordinate change, a first reference speed, a first virtual heading value, a signal-to-noise ratio average of satellite data, and a global positioning system update time of a target user in forming a target user trajectory through a global positioning system, determines first trajectory feature information of the target user trajectory through the obtained first straight line distance difference, first coordinate change, first reference speed, first virtual heading value, signal-to-noise ratio average of satellite data, and global positioning system update time. Then, a step length heading vector, a second straight line distance difference, a second coordinate change, and a second reference speed of the target user trajectory are calculated based on a pedestrian dead reckoning algorithm, and the step length heading vector, second straight line distance difference, second coordinate change, and second reference speed are taken as second trajectory feature information of the target user trajectory.

[0083] The calculation method of the first trajectory feature information and the second trajectory feature information in the embodiment is as follows:

[0084] 1) Extract a step length heading vector provided by a pedestrian dead reckoning algorithm in a target user position point update process obtained by a global positioning system twice continuously wherein, and is a step length heading variable provided by the pedestrian dead reckoning algorithm.

[0085] 2) Extract a first straight line distance difference of a position updated by a global positioning system and a second straight line distance difference of a position updated by a pedestrian dead reckoning algorithm The calculation formula of the first straight line distance difference and the second straight line distance difference is as follows: wherein, and respectively represent a step length heading vector provided by a pedestrian dead reckoning algorithm, represents a position coordinate value output by a global positioning system.

[0086] 3) Extract a first coordinate change of a position updated by a global positioning system and a second coordinate change of a position updated by a pedestrian dead reckoning algorithm wherein, the calculation formula of the first coordinate change and the second coordinate change is as follows: wherein, and respectively represent a step length heading value provided by a pedestrian dead reckoning algorithm, represents a position coordinate value output by a GPS.

[0087] 4) Extract a first reference speed updated by a global positioning system and a second reference speed updated by a pedestrian dead reckoning algorithm wherein the calculation formula of the first reference speed and the second reference speed is: In the formula, and respectively represent the coordinate change of the position updated by the pedestrian dead reckoning algorithm and the GPS, and respectively represent the frequency of the pedestrian dead reckoning algorithm and the GPS.

[0088] 5) Extracting the first virtual heading value updated by the global positioning system and the second virtual heading value updated by the pedestrian dead reckoning algorithm The calculation formula of the first virtual heading value and the second virtual heading value is:

[0089] 6) Extracting the mean value of the signal-to-noise ratio of satellite data obtained by the global positioning system In the formula, SNR η (k) represents the signal-to-noise ratio variable received by each satellite, and a is the number of satellites.

[0090] 7) Extracting the time t required for one update of the global positioning system;

[0091] After obtaining the first trajectory feature information and the second trajectory feature information, the third trajectory feature information can be calculated according to the first trajectory feature information and the second trajectory feature information. The first trajectory feature information, the second trajectory feature information and the third trajectory feature information are taken as the trajectory feature information of the target user trajectory, so as to accurately describe the position point change characteristics in the activity process of the target user.

[0092] In an implementation manner, the step S203 comprises:

[0093] Step S2031, determining a straight line distance residual according to the first straight line distance difference and the second straight line distance difference;

[0094] Step S2032, taking the straight line distance residual as the third trajectory feature information.

[0095] Specifically, the straight line distance residual of the position updated by the global positioning system and the position updated by the pedestrian dead reckoning algorithm is calculated according to the first straight line distance difference in the first trajectory feature information and the second straight line distance difference in the second trajectory feature information , and the calculated straight line distance residual is taken as the third trajectory feature information in the embodiment. The calculation method of the straight line distance residual in the embodiment is:

[0096] As Figure 1As shown, the method further comprises the following steps:

[0097] In step S300, the terrain feature information and the trajectory feature information are input into the hybrid deep learning network to obtain the uncertainty trajectory error corresponding to each position point on the target user trajectory.

[0098] Specifically, to calculate the potential area of the target user trajectory, the embodiment first inputs the terrain feature information and the trajectory feature information into the hybrid deep learning model that has been built to analyze and learn, and obtains the uncertainty trajectory error of the target user. The size of the uncertainty area (potential area) of the target user trajectory is determined according to the uncertainty trajectory error. In the embodiment, by inputting the terrain feature information and the trajectory feature information into the hybrid deep learning model, the time continuity of the error and the feature correlation can be taken into account at the same time, and the accuracy of the uncertainty trajectory error calculation is improved.

[0099] In an implementation manner, the hybrid deep learning network comprises a bidirectional long short-term memory network and a multilayer perceptron.

[0100] Specifically, the hybrid deep learning network in the embodiment is composed of a one-dimensional convolutional neural network, a bidirectional long short-term memory network and a multilayer perceptron, and the structure is as shown in Figure 3 .

[0101] The one-dimensional convolutional neural network is designed in the embodiment, which comprises a convolutional layer, a pooling layer and a flattening layer. The one-dimensional convolutional neural network is used as the first layer of the hybrid deep learning model, is used for learning and extracting the trajectory feature, and comprises determining the format, dimension and specific parameters of the input and output vectors and the neural network connection mode:

[0102]

[0103] In the formula, y j is an output value of the one-dimensional convolutional neural network; x i is an input value, which comprises the terrain feature information and the trajectory feature information; ζ represents a proportion value, that is, a scaling factor of the hybrid deep learning model, k ij and b j respectively represent a connection weight and a bias value, and the connection weight and the bias value in the application are obtained by training the hybrid deep learning model.

[0104] After the one-dimensional convolutional neural network, a bidirectional long short-term memory network model (Bi-LSTM Layer) is constructed as the second layer of the hybrid deep learning model, which is used for learning and training the continuous feature vector:

[0105]

[0106] wherein X t is a feature input, h t-1 represents the input value of the previous node received, h t represents the output value transmitted to the next node, W f , W i , W C , and W o are respective weight values. b f , b i , b C , and b o are respective bias compensation values.

[0107] The full connection layer (Multilayer Perceptron) connected after the bidirectional long short-term memory network model has a multilayer perceptron as the third layer of the hybrid deep learning model, to obtain the output result of the bidirectional long short-term memory network and the estimation result of the final output uncertainty trajectory error wherein y i is the bidirectional long short-term memory network output value, is the final output value (Output Vector) of the uncertainty trajectory error.

[0108] As Figure 1 shown, the method further includes the following steps:

[0109] Step S400, determining a potential area corresponding to the target user trajectory according to the terrain feature information and each uncertainty trajectory error.

[0110] Specifically, the potential area corresponding to the target user trajectory is the approximate range of the real activity of the target user at a position point on the target user trajectory. The size of the potential area corresponding to each position point on the target user trajectory is closely related to the terrain feature information and the uncertainty trajectory error of the position point. Generally speaking, when the slope in the terrain feature information corresponding to a position point on the target user trajectory is steeper, the uncertainty trajectory error is larger, and the potential area corresponding to the position point on the target user trajectory is larger; on the contrary, when the slope in the terrain feature information corresponding to a position point on the target user trajectory is gentler, the uncertainty trajectory error is smaller, and the potential area corresponding to the position point on the target user trajectory is smaller. The embodiment determines the potential area of the target user according to the terrain feature information and the uncertainty trajectory error, comprehensively considers the influence of environmental factors, and improves the accuracy of the potential area estimation.

[0111] In an implementation manner, the step S400 includes:

[0112] Step S401, obtaining a coordinate position corresponding to each position point on the target user trajectory;

[0113] Step S402, constructing a potential region corresponding to each position point according to the coordinate position, the terrain feature information and the uncertainty trajectory error corresponding to each position point;

[0114] Step S403, calculating a union region of the potential regions corresponding to each position point, and taking the union region as the potential region corresponding to the target user trajectory.

[0115] Specifically, the embodiment first obtains a coordinate position corresponding to each position point on the target user trajectory, which is the latitude and longitude coordinate corresponding to the position point and does not include the value of the height direction of the position point on the ground. After obtaining the coordinate position corresponding to each position point, the coordinate position is taken as the center, and the potential region corresponding to the position point is constructed according to the terrain feature information and the uncertainty trajectory error corresponding to the position point. When the potential regions corresponding to all position points on the target user trajectory are constructed, the potential regions corresponding to each position point are merged, and the merged region is taken as the potential region of the target user trajectory. In the embodiment, the target user trajectory is divided into several position points, and the potential region of each position point is determined according to the different terrain environment and uncertainty trajectory error of the position point, so that the method in the embodiment can be applied to various terrains and has good generalization performance.

[0116] In an implementation manner, the step S402 includes:

[0117] Step S4021, when the slope in the terrain feature information corresponding to the position point is not equal to zero, determining that the potential region shape is an ellipse;

[0118] Step S4022, determining an ellipse long semi-axis value according to the terrain feature information, taking the position coordinate of the position point as the center of the ellipse, taking the uncertainty trajectory error corresponding to the position point as the short semi-axis value of the ellipse, and determining the potential region corresponding to the position point.

[0119] Specifically, since the user is in an environment with different terrain feature information, the potential region corresponding to the motion trajectory is different. Therefore, the embodiment first judges the terrain feature information corresponding to the position point on the target user trajectory, when the slope in the terrain feature information corresponding to the position point is not equal to zero (that is, the terrain corresponding to the position point is a slope, not a flat ground), it is determined that the shape of the potential region corresponding to the position point is an ellipse, and the potential region corresponding to the position point is obtained according to the solving method of the ellipse equation.

[0120] The method for solving the potential region corresponding to a position point on the target user's trajectory according to the ellipse variance in the embodiment is: multiplying the slope value of the position point by a preset coefficient to obtain the long semi-axis value of the ellipse, taking the uncertainty trajectory error corresponding to the position point as the short semi-axis of the ellipse, and taking the obtained coordinate position of the position point as the center of the ellipse. The potential region of the position point is solved according to the above values, and the formula is In the formula, (x0, y0) is the center of the ellipse, and η1 and η2 represent the values of the long semi-axis and the short semi-axis respectively.

[0121] In an implementation manner, after the potential region of a single position point is obtained, the union of the potential region of the position point and the potential region of the previous position point can be calculated, and the solving method of the union is:

[0122] 1) solving the intersection points (x1, y1) and (x2, y2) of the two adjacent position points corresponding to the ellipse (potential region);

[0123] 2) calculating the angle value θ of the ellipse to one of the intersection points according to the intersection points and the center of the ellipse of the two ellipses;

[0124] 3) solving the union area of the two adjacent position points corresponding to the ellipse (potential region): and further solving the area of the entire uncertainty region.

[0125] In an implementation manner, the step S402 further includes:

[0126] Step S4023, when the slope in the terrain feature information corresponding to the position point is equal to zero, determining that the shape of the potential region is a circle;

[0127] Step S4024, taking the position coordinate of the position point as the center of the circle, taking the uncertainty trajectory error corresponding to the position point as the radius of the circle, and determining the potential region corresponding to the position point.

[0128] Specifically, when the terrain feature corresponding to a position point on the target user's trajectory is a flat ground, at this time, the slope value of the terrain feature information is zero, it is determined that the shape of the potential region corresponding to the position point is a circle, and the potential region corresponding to the position point is obtained in the form of solving a circle.

[0129] The method for solving the circular potential region in the embodiment is: taking the position coordinate of the position point as the center of the circle, and taking the uncertainty trajectory error corresponding to the position point as the radius of the circle to construct the potential region, and the formula is: (x-a) 2 +(y-b) 2 =r 2 ; in the formula, (a, b) is the position coordinate of the center of the circle, and r is the radius of the circle.

[0130] In an implementation, after obtaining the potential area of a single position point, a union of the potential area of the single position point and the potential area of a previous position point is calculated, and the solution method of the union further includes:

[0131] 1) solving the intersection of two adjacent position points corresponding circles (potential areas): wherein (x0, y0) represents the midpoint of the two intersection points of the circles, (x1, y1) and (x2, y2) represent the intersection point coordinates of the two circular areas (potential areas) respectively.

[0132] 2) solving the area of the union of the two adjacent circular areas (potential areas): and further solving the area of the entire potential area; wherein θ1 and θ2 respectively represent the angle values of the centers of the two circles to the intersection points ， r1 and r2 are the radii of the two circles respectively, and d is the distance between the centers of the two circles.

[0133] In the embodiment, different methods are adopted to construct the potential area corresponding to a single position point on the target user trajectory according to different terrain feature information, the potential area of the entire target user trajectory is obtained by merging the potential areas of the single position points, the time-varying error caused by the global positioning system can be well adapted to while considering the influence of the environment and terrain factors, and the accuracy of the potential area estimation of the target user trajectory is effectively improved.

[0134] Based on the above embodiment, the application further provides a large-range trajectory potential area estimation device considering terrain features, as shown in Figure 4 The device includes:

[0135] A terrain feature acquisition module 01 is configured to acquire terrain feature information corresponding to a region where a target user trajectory is located, wherein the terrain feature information includes the slope, slope direction and slope height of each position point on the target user trajectory.

[0136] A trajectory feature acquisition module 02 is configured to acquire trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm, wherein the trajectory feature information includes information determined based on the global positioning system and the pedestrian dead reckoning algorithm respectively to describe the change of the target user trajectory.

[0137] A trajectory error learning module 03 is configured to input the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain an uncertainty trajectory error corresponding to each position point on the target user trajectory respectively.

[0138] A potential area determination module 04 is configured to determine a potential area corresponding to the target user trajectory according to the terrain feature information and the uncertainty trajectory error.

[0139] Based on the above-mentioned embodiments, the application further provides a terminal, a principle block diagram of which can be shown in the figure. Figure 5 The terminal includes a processor, a memory, a network interface, a display screen connected through a system bus. The processor of the terminal is used to provide computing and control capabilities. The memory of the terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the terminal is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement the large-range trajectory potential area estimation method considering terrain features. The display screen of the terminal can be a liquid crystal display screen or an electronic ink display screen.

[0140] Those skilled in the art can understand that, Figure 5 The principle block diagram shown in the figure is only a block diagram of part of the structure related to the application scheme, and does not constitute a limitation on the terminal to which the application scheme is applied. The specific terminal can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0141] In an implementation manner, the memory of the terminal stores more than one program, and is configured to execute the more than one program by more than one processor, which contains instructions for performing the large-range trajectory potential area estimation method considering terrain features.

[0142] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of each method can be included. Any reference to memory, storage, database or other medium used in each embodiment of the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0143] In summary, the present application discloses a large-range trajectory potential area estimation method and device considering terrain features. The method obtains terrain feature information corresponding to the area where the target user trajectory is located, and obtains trajectory feature information determined based on a global positioning system and a pedestrian dead reckoning algorithm. The terrain feature information and the trajectory feature information are input into a hybrid deep learning network to obtain uncertainty trajectory errors corresponding to each position point on the target user trajectory. The potential area corresponding to the target user trajectory is determined according to the terrain feature information and the uncertainty trajectory errors. The potential area determined according to the terrain feature information and the trajectory feature information considers the complexity of the environment and the error caused by the update of the global positioning system, and solves the problem that the predicted motion trajectory potential area is not accurate due to the neglect of the time-varying measurement error of the global positioning system and the influence of the urban terrain environment in the existing potential area prediction model.

[0144] It should be understood that the application of the present application is not limited to the above examples, and those skilled in the art can make improvements or changes according to the above description, and all these improvements and changes shall belong to the protection scope of the claims of the present application.

Claims

1. A method for estimating a large-scale potential trajectory region that takes into account terrain features, characterized in that, The method includes: Obtain terrain feature information corresponding to the area where the target user's trajectory is located, wherein the terrain feature information includes the slope, aspect, and height of each location point on the target user's trajectory; The trajectory feature information is obtained based on the Global Positioning System and the pedestrian trajectory estimation algorithm, wherein the trajectory feature information includes information used to describe the trajectory changes of the target user, determined based on the Global Positioning System and the pedestrian trajectory estimation algorithm respectively. The terrain feature information and the trajectory feature information are input into a hybrid deep learning network to obtain the uncertainty trajectory error corresponding to each of the location points on the target user's trajectory. Based on the terrain feature information and each of the uncertain trajectory errors, the potential area corresponding to the target user trajectory is determined, including: obtaining the coordinate positions corresponding to each of the location points on the target user trajectory; When the slope in the terrain feature information corresponding to the location point is not equal to zero, the potential area is determined to be an ellipse. The major axis of the ellipse is determined based on the terrain feature information. The location coordinates of the location point are used as the center of the ellipse, and the uncertain trajectory error corresponding to the location point is used as the minor axis of the ellipse to determine the potential area corresponding to the location point. The method for solving the potential area corresponding to a location point on the target user's trajectory based on the ellipse variance is as follows: multiply the slope value of the location point by a preset coefficient to obtain the major axis of the ellipse, use the uncertain trajectory error corresponding to the location point as the minor axis of the ellipse, and use the obtained coordinates of the location point as the center of the ellipse. The potential area of ​​the location point is solved based on the above values, using the following formula: In the formula, (x0, y0) is the center of the ellipse. and Let x and y represent the values ​​of the major and minor semi-axis, respectively. After obtaining the potential region of a single location point, calculate the union of the potential region of that location point and the potential region of the previous location point. This union is obtained by finding the intersection points (x1, y1) and (x2, y2) of the ellipses corresponding to two adjacent location points. Based on the intersection points of the two ellipses and the center of the ellipse, calculate the angle from the ellipse to one of the intersection points. Find the area of ​​the union of the ellipses corresponding to two adjacent points: Solve for the area of ​​the entire uncertainty region; When the slope in the terrain feature information corresponding to the location point is equal to zero, the potential region is determined to be circular. The location coordinates of the location point are used as the center of the circle, and the uncertain trajectory error corresponding to the location point is used as the radius of the circle to determine the potential region corresponding to the location point. The method for solving for a circular potential region is as follows: Using the location coordinates of the location point as the center and the uncertain trajectory error corresponding to the location point as the radius of the circle, the potential region is constructed using the following formula: In the formula, (a, b) are the coordinates of the center of the circle, and r is the radius of the circle. After obtaining the potential region of a single location point, the union of the potential region of that location point and the potential region of the previous location point is calculated. The method for solving this union also includes: solving for the intersection of the circles corresponding to two adjacent location points. In the formula, (x0, y0) represents the midpoint of the two intersection points of the circles, and (x1, y1) and (x2, y2) represent the coordinates of the intersection points of the two circular regions, respectively. The area of ​​the union of two adjacent circular regions is calculated as follows: Solve for the area of ​​the entire potential region; where, and Let r1 and r2 represent the angles from the centers of the two circles to their intersection point, respectively; r1 and r2 are the radii of the two circles, respectively; and d is the distance between the centers of the two circles. Calculate the union of the potential regions corresponding to each of the aforementioned location points, and use the union of the potential regions as the potential regions corresponding to the target user trajectory.

2. The method for estimating a large-scale potential trajectory region considering terrain features according to claim 1, characterized in that, The acquisition of trajectory feature information determined based on the Global Positioning System and pedestrian trajectory estimation algorithm includes: Obtain first trajectory feature information determined based on the global positioning system, wherein the first trajectory feature information includes a first straight-line distance difference, a first coordinate change, a first reference speed, a first virtual heading value, the average signal-to-noise ratio of satellite data, and the global positioning system update time determined based on the global positioning system; Obtain second trajectory feature information determined based on the pedestrian trajectory estimation algorithm, wherein the second trajectory feature information includes step-length heading vector, second straight-line distance difference, second coordinate change and second reference speed determined based on the pedestrian trajectory estimation algorithm; The third trajectory feature information is determined based on the first trajectory feature information and the second trajectory feature information; The trajectory feature information is determined based on the first trajectory feature information, the second trajectory feature information, and the third trajectory feature information.

3. The method for estimating a large-scale potential trajectory region considering terrain features according to claim 2, characterized in that, The step of determining the third trajectory feature information based on the first trajectory feature information and the second trajectory feature information includes: The straight-line distance residual is determined based on the first straight-line distance difference and the second straight-line distance difference; The straight-line distance residual is used as the third trajectory feature information.

4. The method for estimating a large-scale potential trajectory region considering terrain features according to claim 1, characterized in that, The hybrid deep learning network includes a bidirectional long short-term memory network and a multilayer perceptron.

5. A device for estimating a large-scale potential trajectory region that takes into account terrain features, characterized in that, The device includes: The terrain feature acquisition module is used to acquire terrain feature information corresponding to the area where the target user's trajectory is located, wherein the terrain feature information includes the slope, aspect, and height of each location point on the target user's trajectory; The trajectory feature acquisition module is used to acquire trajectory feature information determined based on the global positioning system and the pedestrian trajectory estimation algorithm, wherein the trajectory feature information includes information used to describe the trajectory changes of the target user, determined based on the global positioning system and the pedestrian trajectory estimation algorithm respectively. The trajectory error learning module is used to input the terrain feature information and the trajectory feature information into a hybrid deep learning network to obtain the uncertain trajectory error corresponding to each of the location points on the target user's trajectory. The potential region determination module is used to determine the potential region corresponding to the target user trajectory based on the terrain feature information and each of the uncertain trajectory errors, including: obtaining the coordinate positions corresponding to each of the location points on the target user trajectory; When the slope in the terrain feature information corresponding to the location point is not equal to zero, the potential area is determined to be an ellipse. The major axis of the ellipse is determined based on the terrain feature information. The location coordinates of the location point are used as the center of the ellipse, and the uncertain trajectory error corresponding to the location point is used as the minor axis of the ellipse to determine the potential area corresponding to the location point. The method for solving the potential area corresponding to a location point on the target user's trajectory based on the ellipse variance is as follows: multiply the slope value of the location point by a preset coefficient to obtain the major axis of the ellipse, use the uncertain trajectory error corresponding to the location point as the minor axis of the ellipse, and use the obtained coordinates of the location point as the center of the ellipse. The potential area of ​​the location point is solved based on the above values, using the following formula: In the formula, (x0, y0) is the center of the ellipse. and Let x and y represent the values ​​of the major and minor semi-axis, respectively. After obtaining the potential region of a single location point, calculate the union of the potential region of that location point and the potential region of the previous location point. This union is obtained by finding the intersection points (x1, y1) and (x2, y2) of the ellipses corresponding to two adjacent location points. Based on the intersection points of the two ellipses and the center of the ellipse, calculate the angle from the ellipse to one of the intersection points. Find the area of ​​the union of the ellipses corresponding to two adjacent points: Solve for the area of ​​the entire uncertainty region; When the slope in the terrain feature information corresponding to the location point is equal to zero, the potential region is determined to be circular. The location coordinates of the location point are used as the center of the circle, and the uncertain trajectory error corresponding to the location point is used as the radius of the circle to determine the potential region corresponding to the location point. The method for solving for a circular potential region is as follows: Using the location coordinates of the location point as the center and the uncertain trajectory error corresponding to the location point as the radius of the circle, the potential region is constructed using the following formula: In the formula, (a, b) are the coordinates of the center of the circle, and r is the radius of the circle. After obtaining the potential region of a single location point, the union of the potential region of that location point and the potential region of the previous location point is calculated. The method for solving this union also includes: solving for the intersection of the circles corresponding to two adjacent location points. In the formula, (x0, y0) represents the midpoint of the two intersection points of the circles, and (x1, y1) and (x2, y2) represent the coordinates of the intersection points of the two circular regions, respectively. The area of ​​the union of two adjacent circular regions is calculated as follows: Solve for the area of ​​the entire potential region; where, and Let r1 and r2 represent the angles from the centers of the two circles to their intersection point, respectively; r1 and r2 are the radii of the two circles, respectively; and d is the distance between the centers of the two circles. Calculate the union of the potential regions corresponding to each of the aforementioned location points, and use the union of the potential regions as the potential regions corresponding to the target user trajectory.

6. A terminal, characterized in that, The terminal includes a memory and one or more processors; the memory stores one or more programs; the programs contain instructions for performing the large-scale trajectory potential area estimation method considering terrain features as described in any one of claims 1-4; the processor is used to execute the programs.

7. A computer-readable storage medium storing a plurality of instructions thereon, characterized in that, The instructions are applicable to being loaded and executed by a processor to implement the steps of the large-scale trajectory potential area estimation method considering terrain features as described in any one of claims 1-4.

Images