CN108596941B - Target body motion trajectory prediction method and system based on depth image - Google Patents

Abstract

The invention relates to the technical field of image processing, in particular provides a target body motion track prediction method and a target body motion track prediction system based on a depth image, and aims to solve the technical problem of how to conveniently obtain a target body track with high precision. For this purpose, the target body motion track prediction method can predict a plurality of motion tracks in a segmented manner based on the target body track points in the target body depth image, and further can select the motion track with higher accuracy according to the space distance between the target body track points and the motion tracks so as to obtain the optimal target motion track according to the motion tracks. Meanwhile, the target body motion track prediction system can execute and realize the method.

Description

Target body motion trajectory prediction method and system based on depth image

Technical Field

The invention relates to the technical field of image processing, in particular to a target body motion track prediction method and system based on a depth image.

Background

The badminton robot is an intelligent control system based on technologies such as computers and visual tracking, and can simulate badminton of human beings. The badminton track prediction is an important factor influencing the action accuracy of the badminton robot. Currently, badminton trajectories can be predicted using a physical badminton model or a neural network-based computational model. The badminton trajectory is predicted by adopting the badminton physical model, the operation is simple, the implementation is easy, and the prediction precision is low. The badminton track is predicted by adopting a calculation model based on the neural network, although the prediction precision is higher, the prediction track with higher precision can be obtained only by using training samples with larger orders of magnitude.

Disclosure of Invention

The method aims to solve the technical problem in the prior art, namely how to conveniently acquire the target body track with higher precision. For the purpose, the invention provides a target body motion track prediction method and a target body motion track prediction system based on a depth image.

In a first aspect, the present invention provides a depth image-based target motion trajectory prediction method, including:

acquiring a plurality of continuous target body track points according to a target body depth image acquired in advance;

based on a preset track prediction method, obtaining a first motion track according to 1 st to 3 rd continuous target body track points;

based on the preset track prediction method, obtaining a second motion track according to 2 nd to 4 th continuous target body track points;

based on the preset track prediction method, obtaining a third motion track according to 3 rd to 5 th continuous target body track points;

and correcting the first motion track, the second motion track and the third motion track to obtain a target body motion track.

Further, a preferred technical solution provided by the present invention is:

the preset trajectory prediction method specifically includes:

acquiring the spatial positions of three continuous target body track points, and calculating the initial movement speed and the initial movement direction of a target body according to the acquired spatial positions;

taking a second target body track point of the three continuous target body track points as an initial position point; calculating a target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and according to a method shown in the following formula:

wherein (x)_i,y_i,y_i) The spatial position coordinates of the ith target body track point in the current motion track are obtained; (v)_0x,v_0y,v_0z) As the initial position pointSpatial position coordinates; v is_iThe motion speed of the object at the ith object track point is α_iThe motion direction of the object body at the ith object body track point is taken as the motion direction of the object body; said t is_sThe sampling interval for acquiring the depth image of the target body by the camera device.

Further, a preferred technical solution provided by the present invention is:

the moving speed v of the target body at the ith target body track point_iAnd direction of motion α_iAs shown in the following formula:

wherein, v is_i-1The motion speed of the object at the i-1 th object track point is shown as α_i-1The motion direction of the object body at the ith-1 th object body track point is taken as the motion direction of the object body; k is an air resistance coefficient; the g is the acceleration of gravity.

Further, a preferred technical solution provided by the present invention is:

the step of correcting the first motion track, the second motion track and the third motion track to obtain the motion track of the target body specifically comprises the following steps:

under the condition that the fourth motion track point and the fifth motion track point do not deviate from the first motion track and the fifth motion track point does not deviate from the second motion track, the motion track T of the target body is obtained according to the method shown in the following formula_m：

Under the condition that the fourth motion track point or the fifth motion track point deviates from the first motion track and the fifth motion track point does not deviate from the second motion track, obtaining a target body motion track T according to a method shown in the following formula_m：

In the case where the fourth motion trajectory point or the fifth motion trajectory point deviates from the first motion trajectory, and the fifth motion trajectory point deviates from the second motion trajectory, the target body motion trajectory T_m＝T₃；

Wherein, T is₁、T₂And T₃Respectively a first motion trail, a second motion trail and a third motion trail.

Further, a preferred technical solution provided by the present invention is:

the step of correcting the first motion track, the second motion track and the third motion track to obtain the motion track of the target body comprises the following steps:

calculating the space distance W between the fourth motion track point and the first motion track_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the space distance W between the fifth motion track point and the second motion track_{n+2_1}And W_{n+2_2}；

According to the space distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track: if W_{n+1_1}If the position is more than e, the fourth motion track point deviates from the first motion track;

according to the space distance W_{n+2_1}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the first motion track: if W_{n+2_1}If the position is more than e, the fifth motion track point deviates from the first motion track;

according to the space distance W_{n+2_2}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the second motion track: if W_{n+2_2}And if the second motion track point is larger than e, the fifth motion track point deviates from the second motion track.

In a second aspect, the present invention provides a depth image-based target motion trajectory prediction system, including:

the target body track point acquisition module is configured to acquire a plurality of continuous target body track points according to a pre-acquired target body depth image;

the first motion track acquisition module is configured to obtain a first motion track based on a preset track prediction method according to 1 st to 3 rd target body track points;

the second motion track acquisition module is configured to obtain a second motion track according to 2 nd to 4 th continuous target body track points based on the preset track prediction method;

the third motion track acquisition module is configured to obtain a third motion track according to 3 rd to 5 th continuous target body track points based on the preset track prediction method;

and the target body motion trail acquisition module is configured to correct the first motion trail acquired by the first motion trail acquisition module, the second motion trail acquired by the second motion trail acquisition module and the third motion trail acquired by the third motion trail acquisition module to obtain a target body motion trail.

Further, a preferred technical solution provided by the present invention is:

the system further includes a trajectory prediction module configured to perform operations to enable the system to acquire the first, second, and third motion trajectories:

acquiring the spatial positions of three continuous target body track points, and calculating the initial movement speed and the initial movement direction of a target body according to the acquired spatial positions;

taking a second target body track point of the three continuous target body track points as an initial position point; calculating a target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and according to a method shown in the following formula:

wherein (x)_i,y_i,y_i) The spatial position coordinates of the ith target body track point in the current motion track are obtained; (v)_0x,v_0y,v_0z) The space position coordinates of the initial position point are obtained; v is_iThe motion speed of the object at the ith object track point is α_iThe motion direction of the object body at the ith object body track point is taken as the motion direction of the object body; said t is_sThe sampling interval for acquiring the depth image of the target body by using the camera system.

Further, a preferred technical solution provided by the present invention is:

the trajectory prediction module includes a target body movement speed/direction calculation unit configured to perform the following operations:

calculating the motion speed v of the target body at the ith target body track point according to the method shown in the following formula_iAnd direction of motion α_i：

Wherein, v is_i-1The motion speed of the object at the i-1 th object track point is shown as α_i-1The motion direction of the object body at the ith-1 th object body track point is taken as the motion direction of the object body; k is an air resistance coefficient; the g is the acceleration of gravity.

Further, a preferred technical solution provided by the present invention is:

the target body motion track acquisition module comprises a first track correction unit, a second track correction unit and a third track correction unit;

the first trajectory correction unit is configured to acquire the target body movement trajectory T according to a method shown in the following formula under the condition that the fourth movement trajectory point and the fifth movement trajectory point do not deviate from the first movement trajectory and the fifth movement trajectory point does not deviate from the second movement trajectory_m：

Wherein, T is₁、T₂And T₃Respectively a first motion track, a second motion track and a third motion track;

the second trajectory correction unit is configured to acquire the target body movement trajectory T according to a method shown in the following formula under the condition that the fourth movement trajectory point or the fifth movement trajectory point deviates from the first movement trajectory and the fifth movement trajectory point does not deviate from the second movement trajectory_m：

The third trajectory correction unit is configured to correct the target body movement trajectory T when the fourth movement trajectory point or the fifth movement trajectory point deviates from the first movement trajectory and the fifth movement trajectory point deviates from the second movement trajectory_m＝T₃。

Further, a preferred technical solution provided by the present invention is:

the target body motion track acquisition module further comprises a target body deviation calculation unit, a first target body deviation judgment unit, a second target body deviation judgment unit and a third target body deviation judgment unit;

the target body deviation calculation unit is configured to calculate a spatial distance W between the fourth motion trajectory point and the first motion trajectory_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the space distance W between the fifth motion track point and the second motion track_{n+2_1}And W_{n+2_2}；

The first target body deviation determination unit is configured to determine the first target body deviation according to the spatial distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track: if W_{n+1_1}If the position is more than e, the fourth motion track point deviates from the first motion track;

the second target body deviation determination unit is configured to determine the second target body deviation based on the spatial distance W_{n+2_1}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the first motion track: if W_{n+2_1}If the position is more than e, the fifth motion track point deviates from the first motion track;

the third target body deviation determination unit is configured to determine the spatial distance W_{n+2_2}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the second motion track: if W_{n+2_2}And if the second motion track point is larger than e, the fifth motion track point deviates from the second motion track.

Compared with the closest prior art, the technical scheme at least has the following beneficial effects:

1. the target body motion track prediction method based on the depth image can predict a plurality of motion tracks in a segmented manner based on target body track points in the target body depth image, and further can select motion tracks with high accuracy according to the space distance between the target body track points and the motion tracks so as to obtain the optimal target motion tracks according to the motion tracks.

2. According to the target body motion track prediction method based on the depth image, the threshold value of the space distance is set according to the precision of the camera device, so that whether track points in the current track deviate from the previous track or not can be accurately judged.

3. The target body motion trail prediction method based on the depth image adopts three continuous target body track points to predict a section of motion trail, and can obtain the optimal target body motion trail and the optimal target body floor point according to the three sections of continuous motion trails.

Drawings

FIG. 1 is a schematic diagram illustrating the main steps of a depth image-based target motion trajectory prediction method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram illustrating the main steps of another depth image-based object motion trajectory prediction system according to an embodiment of the present invention;

fig. 3 is a schematic structural diagram of a depth image-based target motion trajectory prediction system according to an embodiment of the present invention.

Detailed Description

Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and are not intended to limit the scope of the present invention.

Referring to fig. 1, fig. 1 schematically illustrates the main steps of a depth image-based target motion trajectory prediction method in this embodiment. As shown in fig. 1, in this embodiment, the motion trajectory of the target body can be predicted according to the following steps:

step S101: and acquiring a plurality of continuous target body track points according to the pre-acquired target body depth image. Specifically, in this embodiment, a camera device, such as a depth camera, may be used to obtain a depth image during the movement of the target, and a conventional depth image processing method may be used to obtain the target track point. For example, in the embodiment, a Depth bases-D2D program disclosed by Kinect for Windows v2.0 SDK may be adopted to obtain the target body track point.

Step S102: and obtaining a first motion track according to the 1 st to 3 rd target body track points based on a preset track prediction method.

Specifically, the preset trajectory prediction method in this embodiment includes the following steps:

firstly, the spatial positions of three continuous target body track points are obtained, and the initial movement speed and the initial movement direction of the target body are calculated according to the obtained spatial positions.

Secondly, taking a second target body track point of the three continuous target body track points as an initial position point; calculating a target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and according to a method shown in the following formula (1):

the meaning of each parameter in the formula (1) is as follows:

(x_i,y_i,y_i) And the spatial position coordinates of the ith target body track point in the current motion track are obtained. (v)_0x,v_0y,v_0z) Is the spatial position coordinates of the initial position point. v. of_iα is the moving speed of the object at the ith object track point_iThe moving direction of the target body at the ith target body track point is shown. t is t_sThe sampling interval for acquiring the depth image of the target body by the camera device.

In this embodiment, the moving speed v of the object at the ith object track point_iAnd direction of motion α_iAs shown in the following formula (2):

the meaning of each parameter in the formula (2) is as follows:

v_i-1α is the moving speed of the object body at the i-1 th object body track point_i-1The moving direction of the target body at the (i-1) th target body track point is shown. k is an air resistance coefficient. g is the acceleration of gravity.

Further, based on the preset trajectory prediction method, the target body motion trajectory prediction method shown in fig. 1 may obtain the first motion trajectory according to the following steps:

step S1021: obtaining 1 st to 3 rd continuous target body track points P₁、P₂And P₃In space, i.e.

And

step S1022: tracing the target body with a point P₂As an initial position point, an initial movement velocity (v) of the target body in the first movement track can be calculated according to the following formulas (3) and (4)_0x,v_0y,v_0z) And an initial direction of motion α₀：

Step S1023: and calculating other target body track points in the first motion track according to the initial position point, the initial motion speed and the initial motion direction and the method shown in the formula (1) to further obtain the first motion track.

Step S103: and obtaining a second motion track according to 2 nd to 4 th continuous target body track points based on a preset track prediction method. In this embodiment, the method for acquiring the second motion trajectory is the same as the method for acquiring the first motion trajectory, and specifically includes the following steps:

step S1031: obtaining track points P of 2 nd to 4 th continuous object bodies₂、P₃And P₄In space, i.e.

And

step S1032: tracing the target body with a point P₃Calculating to obtain the initial movement velocity (v) of the target body in the second movement track as the initial position point according to the methods described in the formulas (3) and (4)_0x,v_0y,v_0z) And an initial direction of motion α₀。

Step S1033: and calculating other target body track points in the second motion track according to the initial position point, the initial motion speed and the initial motion direction and the method shown in the formula (1) to further obtain the second motion track.

Step S104: and obtaining a third motion track according to 3 rd to 5 th continuous target body track points based on a preset track prediction method.

Step S1041: obtainTaking the track points P of the 3 rd to 5 th continuous object bodies₃、P₄And P₅In space, i.e.

And

step S1032: tracing the target body with a point P₄Calculating to obtain the initial movement velocity (v) of the target body in the third movement track as the initial position point according to the methods described in the formulas (3) and (4)_0x,v_0y,v_0z) And an initial direction of motion α₀。

Step S1033: and calculating other target body track points in the third motion track according to the initial position point, the initial motion speed and the initial motion direction and the method shown in the formula (1) to further obtain the third motion track.

Step S105: and correcting the first motion trail, the second motion trail and the third motion trail to obtain a motion trail of the target body. Specifically, in this embodiment, the movement locus of the target body may be obtained as follows:

step S1051: judging whether the track point of the target body deviates from the motion track:

firstly, the space distance W between the fourth motion locus point and the first motion locus is calculated_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the second motion track respectively_{n+2_1}And W_{n+2_2}。

Secondly, according to the spatial distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track; according to the spatial distance W_{n+2_1}And a preset threshold value e, judging whether the fifth motion track point deviates from the first motion track; according to the spatial distance W_{n+2_2}And a preset threshold value e, and judging whether the fifth motion track point deviates from the second motion track.

If W_{n+1_1}If the fourth motion track point deviates from the first motion track;

if W_{n+2_1}If the position is more than e, the fifth motion track point deviates from the first motion track;

if W_{n+2_2}And if the position is more than e, the fifth motion track point deviates from the second motion track.

Step S1052: obtaining the movement track T of the target body according to the judgment result_m：

1. Under the condition that the fourth motion track point and the fifth motion track point do not deviate from the first motion track and the fifth motion track point does not deviate from the second motion track, the target motion track T can be obtained according to the method shown in the following formula (5)_m：

The meaning of each parameter in the formula (5) is as follows: t is₁、T₂And T₃Respectively a first motion trail, a second motion trail and a third motion trail.

2. And under the condition that the fourth motion track point or the fifth motion track point deviates from the first motion track and the fifth motion track point does not deviate from the second motion track, acquiring the motion track T of the target body according to the method shown in the following formula (6)_m：

3. In the case where the fourth motion trajectory point or the fifth motion trajectory point deviates from the first motion trajectory, and the fifth motion trajectory point deviates from the second motion trajectory, the target body motion trajectory T_m＝T₃。

Referring to fig. 2, fig. 2 schematically illustrates the main steps of another depth image-based target motion trajectory prediction method in this embodiment. As shown in fig. 2, in this embodiment, the motion trajectory of the target body can be predicted according to the following steps:

step S201: and acquiring a target body depth image.

Step S202: and acquiring the spatial positions of the 1 st to 3 rd continuous target body track points.

Step S203: and calculating the initial movement speed and the initial speed direction of the target body in the current movement track. Specifically, in this embodiment, the initial movement velocity (v) of the target body in the current movement track can be calculated and obtained according to the methods shown in the formulas (3) and (4)_0x,v_0y,v_0z) And an initial direction of motion α₀。

Step S204: and calculating other target body track points in the current motion track according to the initial position point, the initial motion speed and the initial speed direction. Specifically, in this embodiment, other target trajectory points in the current motion trajectory may be calculated according to the method shown in formula (1).

Step S205: and forming a first motion trail according to the 1 st to 3 rd continuous target body track points and other corresponding target body track points.

Step S206: and acquiring the spatial position of the track point of the 4 th object body, and repeatedly executing the steps S203-S204 to obtain a second motion track.

Step S207: and acquiring the spatial position of the track point of the 5 th object body, and repeatedly executing the steps S203-S204 to obtain a third motion track.

Step S208: and judging whether the 4 th target body track point or the 5 th target body track point deviates from the first motion track. Specifically, if the 4 th object trajectory point or the 5 th object trajectory point deviates from the first motion trajectory, the process goes to step S209. If neither the 4 th object trajectory point nor the 5 th object trajectory point deviates from the first motion trajectory, the process goes to step S213.

Step S209: the first motion profile is discarded.

Step S210: and judging whether the 5 th target body track point deviates from the second motion track. Specifically, if the 5 th object trajectory point does not deviate from the second motion trajectory, the process goes to step S211. If the 5 th object trajectory point deviates from the second motion trajectory, the second motion trajectory is abandoned, and the process goes to step S212.

Step S211: and obtaining the motion trail of the target body according to the second motion trail and the third motion trail. Specifically, in this embodiment, the target body motion trajectory may be obtained according to the method shown in formula (6).

Step S212: and obtaining the motion trail of the target body according to the third motion trail. Specifically, in this embodiment, the target body movement locus is a third movement locus.

Step S213: and judging whether the 5 th target body track point deviates from the second motion track. Specifically, if the 5 th object trajectory point deviates from the second motion trajectory, go to step S214. If the 5 th object trajectory point does not deviate from the second motion trajectory, go to step S215.

Step S214: the second motion profile is discarded.

Step S215: and obtaining the motion trail of the target body according to the first, second and third motion trails. Specifically, in this embodiment, the target body motion trajectory may be obtained according to the method shown in formula (5).

Although the foregoing embodiments describe the steps in the above sequential order, those skilled in the art will understand that, in order to achieve the effect of the present embodiments, the steps may not be executed in such an order, and may be executed simultaneously (in parallel) or in an inverse order, and these simple variations are within the scope of the present invention.

Based on the same technical concept as the method embodiment, the embodiment of the invention also provides a target body motion track prediction system based on the depth image. The depth image-based target body motion trajectory prediction system is specifically described below with reference to the drawings.

Referring to fig. 2, fig. 2 schematically shows the main structure of a depth image-based target motion trajectory prediction system in the present embodiment. As shown in fig. 2, the depth image-based target motion trajectory prediction system in this embodiment may include a target trajectory point obtaining module 11, a first motion trajectory obtaining module 12, a second motion trajectory obtaining module 13, a third motion trajectory obtaining module 14, and a target motion trajectory obtaining module 15. Specifically, in this embodiment, the object trajectory point obtaining module 11 may be configured to obtain a plurality of continuous object trajectory points according to a pre-obtained object depth image. The first motion trail obtaining module 12 may be configured to obtain a first motion trail based on a preset trail prediction method according to 1 st to 3 rd target body track points. The second motion trajectory acquisition module 13 may be configured to obtain a second motion trajectory based on a preset trajectory prediction method according to 2 nd to 4 th continuous target body trajectory points. The third motion trajectory acquisition module 14 may be configured to obtain a third motion trajectory according to 3 rd to 5 th continuous target body trajectory points based on a preset trajectory prediction method. The target body movement track obtaining module 15 may be configured to correct the first movement track obtained by the first movement track obtaining module 12, the second movement track obtained by the second movement track obtaining module 13, and the third movement track obtained by the third movement track obtaining module 13, so as to obtain the target body movement track.

Further, in this embodiment, the target trajectory prediction system shown in fig. 2 may further include a trajectory prediction module configured to perform the following operations, so that the system can acquire the first motion trajectory, the second motion trajectory, and the third motion trajectory:

firstly, the spatial positions of three continuous target body track points are obtained, and the initial movement speed and the initial movement direction of the target body are calculated according to the obtained spatial positions.

Secondly, taking a second target body track point of the three continuous target body track points as an initial position point; and calculating the target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and the method shown in the formula (1).

Further, the trajectory prediction module in this embodiment may include an object movement velocity/direction calculation unit configured to calculate the movement velocity v of the object at the ith object trajectory point according to the method shown in formula (2)_iAnd direction of motion α_i。

Further, in this embodiment, the target motion trajectory acquisition module 15 may include a first trajectory correction unit, a second trajectory correction unit, and a third trajectory correction unit.

Specifically, the first trajectory correction unit may be configured to acquire the target body movement trajectory T according to the method shown in formula (5) when the fourth movement trajectory point and the fifth movement trajectory point do not deviate from the first movement trajectory and the fifth movement trajectory point does not deviate from the second movement trajectory_m. The second trajectory correction unit may be configured to acquire the target body movement trajectory T according to the method shown in formula (6) in a case where the fourth movement trajectory point or the fifth movement trajectory point deviates from the first movement trajectory and the fifth movement trajectory point does not deviate from the second movement trajectory_m. The third trajectory correction unit may be configured such that the target body movement trajectory T is in a case where the fourth movement trajectory point or the fifth movement trajectory point deviates from the first movement trajectory, and the fifth movement trajectory point deviates from the second movement trajectory_m＝T₃。

Further, in this embodiment, the target body movement trajectory acquisition module 15 may further include a target body deviation calculation unit, a first target body deviation determination unit, a second target body deviation determination unit, and a third target body deviation determination unit.

Specifically, the target body deviation calculation unit may be configured to calculate a spatial distance W between the fourth motion trajectory point and the first motion trajectory_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the second motion track respectively_{n+2_1}And W_{n+2_2}. The first target body deviation determination unit may be configured to determine the first target body deviation in accordance with the spatial distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track: if W_{n+1_1}And if the position is more than e, the fourth motion track point deviates from the first motion track. The second target body deviation determination unit may be configured to determine the second target body deviation in accordance with the spatial distance W_{n+2_1}And a preset threshold value e, judging whether the fifth motion track point deviates from the first motion track: if W_{n+2_1}And e, deviating the fifth motion track point from the first motion track. The third target body deviation determination unit may be configured to determine the distance W in space_{n+2_2}And a preset threshold value e, judging whether the fifth motion track point deviates from the second motion track: if it isW_{n+2_2}And if the position is more than e, the fifth motion track point deviates from the second motion track.

The above-mentioned target body motion trajectory prediction system embodiment based on the depth image may be used to implement the above-mentioned target body motion trajectory prediction method embodiment based on the depth image, and the technical principle, the solved technical problems, and the generated technical effects are similar, and it can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process and the related description of the above-mentioned target body motion trajectory prediction system based on the depth image may refer to the corresponding process in the above-mentioned target body motion trajectory prediction method embodiment based on the depth image, and are not described herein again.

Those skilled in the art will appreciate that the above-described depth image-based object motion trajectory prediction system also includes some other well-known structures, such as processors, memories, etc., wherein the memories include, but are not limited to, random access memory, flash memory, read only memory, programmable read only memory, volatile memory, non-volatile memory, serial memory, parallel memory or registers, etc., and the processors include, but are not limited to, CPLD/FPGA, DSP, ARM processor, MIPS processor, etc., and these well-known structures are not shown in fig. 2 in order to unnecessarily obscure embodiments of the present disclosure.

It should be understood that the number of individual modules in fig. 2 is merely illustrative. The number of modules may be any according to actual needs.

Those skilled in the art will appreciate that the modules in the system in an embodiment may be adaptively changed and arranged in one or more systems different from the embodiment. The modules or units in the embodiments may be combined into one module or unit, and furthermore, they may be divided into a plurality of sub-modules or sub-units. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements.

Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the claims of the present invention, any of the claimed embodiments may be used in any combination.

So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.

Claims (8)

1. A target body motion track prediction method based on a depth image is characterized by comprising the following steps:

acquiring a plurality of continuous target body track points according to a target body depth image acquired in advance;

based on a preset track prediction method, obtaining a first motion track according to 1 st to 3 rd continuous target body track points;

based on the preset track prediction method, obtaining a second motion track according to 2 nd to 4 th continuous target body track points;

based on the preset track prediction method, obtaining a third motion track according to 3 rd to 5 th continuous target body track points;

correcting the first motion track, the second motion track and the third motion track to obtain a motion track of a target body;

the preset trajectory prediction method specifically includes:

acquiring the spatial positions of three continuous target body track points, and calculating the initial movement speed and the initial movement direction of a target body according to the acquired spatial positions;

taking a second target body track point of the three continuous target body track points as an initial position point; calculating a target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and according to a method shown in the following formula:

wherein (x)_i,y_i,y_i) The spatial position coordinates of the ith target body track point in the current motion track are obtained; (v)_0x,v_0y,v_0z) The space position coordinates of the initial position point are obtained; v is_iThe motion speed of the object at the ith object track point is α_iThe motion direction of the object body at the ith object body track point is taken as the motion direction of the object body; said t is_sThe sampling interval for acquiring the depth image of the target body by the camera device.

2. The depth image-based target motion trail prediction method according to claim 1, wherein the motion velocity v of the target at the ith target track point_iAnd direction of motion α_iAs shown in the following formula:

wherein, v is_i-1The motion speed of the object at the i-1 th object track point is shown as α_i-1The motion direction of the object body at the ith-1 th object body track point is taken as the motion direction of the object body; k is an air resistance coefficient; the g is the acceleration of gravity.

3. The depth-image-based target body motion trajectory prediction method according to claim 1 or 2, wherein the step of modifying the first motion trajectory, the second motion trajectory, and the third motion trajectory to obtain the target body motion trajectory specifically includes:

under the condition that the fourth motion track point and the fifth motion track point do not deviate from the first motion track and the fifth motion track point does not deviate from the second motion track, the motion track T of the target body is obtained according to the method shown in the following formula_m：

Under the condition that the fourth motion track point or the fifth motion track point deviates from the first motion track and the fifth motion track point does not deviate from the second motion track, obtaining a target body motion track T according to a method shown in the following formula_m：

In the case where the fourth motion trajectory point or the fifth motion trajectory point deviates from the first motion trajectory, and the fifth motion trajectory point deviates from the second motion trajectory, the target body motion trajectory T_m＝T₃；

Wherein, T is₁、T₂And T₃Respectively a first motion trail, a second motion trail and a third motion trail.

4. The method for predicting the target body motion trail based on the depth image according to claim 3, wherein the step of modifying the first motion trail, the second motion trail and the third motion trail to obtain the target body motion trail comprises:

calculating the space distance W between the fourth motion track point and the first motion track_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the space distance W between the fifth motion track point and the second motion track_{n+2_1}And W_{n+2_2}；

According to the space distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track: if W_{n+1_1}If the position is more than e, the fourth motion track point deviates from the first motion track;

according to the space distance W_{n+2_1}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the first motion track: if W_{n+2_1}If the position is more than e, the fifth motion track point deviates from the first motion track;

according to the space distance W_{n+2_2}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the second motion track: if W_{n+2_2}And if the second motion track point is larger than e, the fifth motion track point deviates from the second motion track.

5. A depth image-based target body motion trajectory prediction system, comprising:

the target body track point acquisition module is configured to acquire a plurality of continuous target body track points according to a pre-acquired target body depth image;

the first motion track acquisition module is configured to obtain a first motion track based on a preset track prediction method according to 1 st to 3 rd target body track points;

the second motion track acquisition module is configured to obtain a second motion track according to 2 nd to 4 th continuous target body track points based on the preset track prediction method;

the third motion track acquisition module is configured to obtain a third motion track according to 3 rd to 5 th continuous target body track points based on the preset track prediction method;

a target body movement track acquisition module configured to correct the first movement track acquired by the first movement track acquisition module, the second movement track acquired by the second movement track acquisition module, and the third movement track acquired by the third movement track acquisition module to obtain a target body movement track;

the system further includes a trajectory prediction module configured to perform operations to enable the system to acquire the first, second, and third motion trajectories:

acquiring the spatial positions of three continuous target body track points, and calculating the initial movement speed and the initial movement direction of a target body according to the acquired spatial positions;

taking a second target body track point of the three continuous target body track points as an initial position point; calculating a target body track point in the current motion track according to the initial position point, the initial motion speed and the initial motion direction and according to a method shown in the following formula:

wherein (x)_i,y_i,y_i) The spatial position coordinates of the ith target body track point in the current motion track are obtained; (v)_0x,v_0y,v_0z) The space position coordinates of the initial position point are obtained; v is_iThe motion speed of the object at the ith object track point is α_iThe motion direction of the object body at the ith object body track point is taken as the motion direction of the object body; said t is_sFor obtaining purpose by camera systemSampling interval of the object depth image.

6. The depth-image-based target body movement trajectory prediction system according to claim 5, wherein the trajectory prediction module includes a target body movement speed/direction calculation unit configured to perform operations of:

calculating the motion speed v of the target body at the ith target body track point according to the method shown in the following formula_iAnd direction of motion α_i：

Wherein, v is_i-1The motion speed of the object at the i-1 th object track point is shown as α_i-1The motion direction of the object body at the ith-1 th object body track point is taken as the motion direction of the object body; k is an air resistance coefficient; the g is the acceleration of gravity.

7. The depth-image-based target motion trajectory prediction system according to claim 5 or 6, wherein the target motion trajectory acquisition module includes a first trajectory correction unit, a second trajectory correction unit, and a third trajectory correction unit;

the first trajectory correction unit is configured to acquire the target body movement trajectory T according to a method shown in the following formula under the condition that the fourth movement trajectory point and the fifth movement trajectory point do not deviate from the first movement trajectory and the fifth movement trajectory point does not deviate from the second movement trajectory_m：

Wherein, T is₁、T₂And T₃Respectively a first motion track, a second motion track and a third motion track;

the second track correction unit is configured to correct the fourth motion trackAnd under the condition that the point or the fifth motion track point deviates from the first motion track and the fifth motion track point does not deviate from the second motion track, acquiring a target body motion track T according to a method shown in the following formula_m：

The third trajectory correction unit is configured to correct the target body movement trajectory T when the fourth movement trajectory point or the fifth movement trajectory point deviates from the first movement trajectory and the fifth movement trajectory point deviates from the second movement trajectory_m＝T₃。

8. The depth-image-based target body movement track prediction system according to claim 7, wherein the target body movement track acquisition module further includes a target body deviation calculation unit, a first target body deviation determination unit, a second target body deviation determination unit, and a third target body deviation determination unit;

the target body deviation calculation unit is configured to calculate a spatial distance W between the fourth motion trajectory point and the first motion trajectory_{n+1_1}Calculating the space distance W between the fifth motion track point and the first motion track and the space distance W between the fifth motion track point and the second motion track_{n+2_1}And W_{n+2_2}；

The first target body deviation determination unit is configured to determine the first target body deviation according to the spatial distance W_{n+1_1}And a preset threshold value e, judging whether the fourth motion track point deviates from the first motion track: if W_{n+1_1}If the position is more than e, the fourth motion track point deviates from the first motion track;

the second target body deviation determination unit is configured to determine the second target body deviation based on the spatial distance W_{n+2_1}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the first motion track: if W_{n+2_1}If the position is more than e, the fifth motion track point deviates from the first motion track;

the third target body deviation determination unit is configured to determine the spatial distance W_{n+2_2}And the preset threshold value e is used for judging whether the fifth motion track point deviates from the second motion track: if W_{n+2_2}And if the second motion track point is larger than e, the fifth motion track point deviates from the second motion track.

Images