CN112085762B - Target position prediction method based on curvature radius and storage medium - Google Patents
Target position prediction method based on curvature radius and storage medium Download PDFInfo
- Publication number
- CN112085762B CN112085762B CN201910514119.3A CN201910514119A CN112085762B CN 112085762 B CN112085762 B CN 112085762B CN 201910514119 A CN201910514119 A CN 201910514119A CN 112085762 B CN112085762 B CN 112085762B
- Authority
- CN
- China
- Prior art keywords
- calculating
- point
- plane
- target
- projection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a target position prediction method based on curvature radius and a storage medium, wherein the method comprises the following steps: establishing a three-dimensional rectangular coordinate system; calculating coordinates of a position center point of the target in three frames of RGB-D images respectively; obtaining a first plane according to the three position center points and calculating an equation of the first plane; projecting the three position center points to an XOY plane; fitting three projection points by using a circular equation or a parabolic equation, and calculating the curvature radius of the motion curve at the last projection point; calculating an instantaneous speed from the radius of curvature; calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed; calculating a plane predicted point of the target on the motion curve according to the coordinate and the displacement of the last projection point; and making a perpendicular line of the XOY plane through the plane prediction point, and calculating the coordinate of the intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target. The invention can predict the position of the object in the three-dimensional space.
Description
Technical Field
The present invention relates to the field of target position prediction technologies, and in particular, to a target position prediction method based on a radius of curvature and a storage medium.
Background
The target tracking and predicting algorithm is applied to a plurality of intelligent monitoring devices, such as intelligent security cameras, unmanned aerial vehicles, robots and the like. The traditional camera mainly acquires an RGB image of a target, and performs target identification, tracking and position prediction on the RGB image. With the continuous development of camera technology, some depth cameras using structured light are available on the market, and besides the conventional RGB lens, the structured light is also used to obtain depth information of an object. The latest technology can utilize structured light to carry out three-dimensional modeling on objects and spaces, and splice a plurality of point cloud pictures to form a complete three-dimensional point cloud space.
At present, an object tracking task and a position predicting task in a three-dimensional point cloud space have no specific algorithm, and a traditional means shift algorithm can be applied to object tracking in the three-dimensional space, but the accuracy is insufficient, and particularly, when the object tracking process is blocked by other objects, the tracking is easy to be misplaced, and at the moment, the position is required to be predicted for position verification.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: provided are a target position prediction method based on a radius of curvature, which can predict the position of a target object in a three-dimensional space in the next frame, and which has high prediction accuracy.
In order to solve the technical problems, the invention adopts the following technical scheme: a target position prediction method based on a radius of curvature, comprising:
acquiring an RGB image, and establishing a three-dimensional rectangular coordinate system according to the RGB image;
calculating coordinates of a position center point of the target in three frames of RGB-D images respectively;
determining a plane according to the three position center points to obtain a first plane, and calculating an equation of the first plane;
projecting the three position center points to an XOY plane to obtain three projection points, and calculating coordinates of the three projection points;
fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point;
calculating to obtain the instantaneous speed of the target at the last projection point according to the curvature radius;
calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed;
calculating a plane prediction point of the target on the motion curve according to the coordinate of the last projection point and the displacement in the X-axis direction;
and making a perpendicular line of the XOY plane through the plane prediction point, and calculating coordinates of an intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target.
The invention also relates to a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out the steps as described above.
The invention has the beneficial effects that: fitting a curve in a three-dimensional space according to the position center point of the target object in the three-frame depth image, calculating the curvature radius of the position of the point on the curve, calculating the instantaneous speed of the target in the third-frame depth image, and predicting the position of the target object in the next-frame depth image by using the obtained instantaneous speed; the space curve is projected onto the plane by adopting a projection method, so that the calculation difficulty can be reduced, and the calculation efficiency can be improved; by utilizing curve fitting motion trail and utilizing curvature radius to calculate instantaneous speed, the calculated result error is ensured to be smaller, thereby ensuring the accuracy of subsequent position prediction. The method can predict the position of the object in the three-dimensional space, has high accuracy, does not need to model the motion track of the target object in advance at one time, and can be used as an auxiliary means for object tracking task in the three-dimensional space for position verification.
Drawings
FIG. 1 is a flow chart of a method for predicting a target location based on a radius of curvature according to the present invention;
fig. 2 is a flowchart of a method according to a first embodiment of the invention.
Detailed Description
In order to describe the technical contents, the achieved objects and effects of the present invention in detail, the following description will be made with reference to the embodiments in conjunction with the accompanying drawings.
The most critical concept of the invention is as follows: and (3) utilizing the curve to fit the motion track, utilizing the curvature radius to calculate the instantaneous speed, utilizing the instantaneous speed and the frame interval time to calculate the displacement, and utilizing the displacement to predict the position of the next frame.
Referring to fig. 1, a target position prediction method based on a curvature radius includes:
acquiring an RGB image, and establishing a three-dimensional rectangular coordinate system according to the RGB image;
calculating coordinates of a position center point of the target in three frames of RGB-D images respectively;
determining a plane according to the three position center points to obtain a first plane, and calculating an equation of the first plane;
projecting the three position center points to an XOY plane to obtain three projection points, and calculating coordinates of the three projection points;
fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point;
calculating to obtain the instantaneous speed of the target at the last projection point according to the curvature radius;
calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed;
calculating a plane prediction point of the target on the motion curve according to the coordinate of the last projection point and the displacement in the X-axis direction;
and making a perpendicular line of the XOY plane through the plane prediction point, and calculating coordinates of an intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target.
From the above description, the beneficial effects of the invention are as follows: the position prediction can be carried out on the object in the three-dimensional space, and the accuracy is high.
Further, the establishing a three-dimensional rectangular coordinate system according to the RGB image specifically includes:
and establishing a three-dimensional rectangular coordinate system by taking the upper left corner of the RGB image as an origin, taking a ray passing through the origin and vertically downwards as a Y-axis positive direction, taking a ray passing through the origin and horizontally rightwards as a Z-axis positive direction, and taking a ray passing through the origin and vertical to the inward direction of the RGB image as the Z-axis positive direction.
Further, after the three location center points are projected to the XOY plane to obtain three projection points and coordinates of the three projection points are calculated, the method further includes:
judging whether the three projection points are positioned on the same straight line or not;
if yes, transforming the projection surface;
and if not, executing the step of fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point.
From the above description, when three projection points appear to be collinear, the projection points are made non-collinear by transforming the projection plane.
Further, the fitting of the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the radius of curvature of the motion curve at the last projection point is specifically as follows:
judging whether the X coordinates of two projection points in the three projection points are equal in value or not;
if the motion curve exists, fitting the three projection points through a circular equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point;
if the motion curve does not exist, fitting the three projection points by using a parabolic equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point.
From the above description, it can be seen that when the X coordinate values of the two projection points are equal, the parabolic equation is not solved, and thus a circular equation is used for fitting.
Further, according to the curvature radius, calculating to obtain the instantaneous speed of the target at the last projection point specifically includes:
calculating an included angle between the normal direction of the motion curve at the last projection point and the gravity direction;
calculating the instantaneous speed of the target at the last projection point according to a first formula, wherein the first formula is mv 2 R=mgcos θ, m is the target mass, v is the instantaneous velocity, r is the radius of curvature, g is the gravitational acceleration, θ is the included angle.
From the above description, the instantaneous velocity is calculated by using the centripetal force formula of the curve motion and the principle that the centripetal force formula is the component of the gravity of the object in the curve normal direction.
Further, the calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed specifically includes:
calculating an X-axis component of the instantaneous velocity by orthogonal decomposition;
and calculating the displacement of the target in the X-axis direction according to the frame interval time and the X-axis component.
As is apparent from the above description, the displacement amount of the target between the adjacent two frame images is calculated using the instantaneous speed and the frame interval time.
Further, the perpendicular to the XOY plane is drawn through the plane prediction point, and coordinates of an intersection point of the perpendicular and the first plane are calculated, so that a predicted position point of the target is obtained specifically as follows:
making a perpendicular line of the XOY plane through the plane prediction point, and calculating an equation of the perpendicular line;
and calculating coordinates of an intersection point of the vertical line and the first plane according to the equation of the vertical line and the equation of the first plane to obtain a predicted position point of the target.
From the above description, it is known that, theoretically, a curve in the XOY plane is translated in the positive Z-axis direction to form a curve plane, and a curve where the curve plane intersects with the first plane Q is a motion curve of the object in the three-dimensional space.
The invention also proposes a computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out the steps as described above.
Example 1
Referring to fig. 2, a first embodiment of the invention is as follows: the method is mainly used for tracking and predicting the falling objects in life, such as predicting the track of objects such as basketball, volleyball, table tennis and the like (subsequent hit rate prediction can be performed), and comprises the following steps:
s1: acquiring an RGB image, and establishing a three-dimensional rectangular coordinate system according to the RGB image; specifically, in this embodiment, a three-dimensional rectangular coordinate system is established with the upper left corner of the RGB image as the origin, the ray passing through the origin and vertically downward as the positive Y-axis direction, the ray passing through the origin and horizontally rightward as the positive X-axis direction, and the ray passing through the origin and perpendicular to the inward direction of the RGB image as the positive Z-axis direction.
S2: calculating coordinates of a position center point of the target in three frames of RGB-D images respectively; i.e. each of three RGB-D images (depth images) gives a position center point of the object, i.e. three position center points in total, assuming P in turn 1 、P 2 、P 3 。
Wherein, three frames of RGB-D images can be three frames continuously or discontinuous; the purpose of the method is to predict the position point P of the target in the next frame RGB-D image of the last frame of the three frames RGB-D images 4 . The time interval between each frame is required to be known.
Further, in this embodiment, all the point coordinates of the target may be obtained from the three-dimensional point cloud corresponding to the three frames of RGB-D images, and then the average value of all the point coordinates may be calculated as the coordinates of the position center point of the target.
S3: judging whether the three position center points are on the same straight line, if so, exiting, namely not predicting the straight line movement, and if not, executing the step S4, wherein the method only predicts the position of the curve movement.
S4: and determining a plane according to the three position center points to obtain a first plane, and calculating an equation of the first plane. According to the principle of determining a plane by three points, a first plane Q can be obtained according to three non-collinear position center points, and the equation is a 1 X+b 1 Y+c 1 Z+d 1 =0, wherein by substituting three location center points P 1 、P 2 And P 3 Coordinates of (a) to obtain a 1 、b 1 、c 1 And d 1 。
S5: and projecting the three central points to an XOY plane to obtain three projection points, and calculating coordinates of the three projection points. Assume three location center points P 1 、P 2 、P 3 The corresponding projection points are P respectively 1 ’、P 2 ’、P 3 ' let P in principle 1 、P 2 、P 3 The value of the Z axis in the coordinate value of (2) is zero, and P can be obtained 1 ’、P 2 ’、P 3 ' coordinates.
S6: and judging whether the three projection points are on the same straight line, if so, executing the step S7, and if not, executing the step S8.
S7: the projection plane is transformed, such as by projecting the three location center points to the ZOY plane. Since the above steps have guaranteed three location center points P 1 、P 2 、P 3 Are not collinear in space, so that when three projection points appear collinear, the projection points can be made non-collinear by transforming the projection plane. Further, for format unification, the present embodiment realizes the transformation of the projection plane by exchanging the X-axis and the Z-axis of the three-dimensional rectangular coordinate system, and recalculates the three-position center point P 1 、P 2 、P 3 The equation of the first plane Q and the coordinates of (c) is returned to perform step S2.
Further, the coordinates of the new three-position center point can be obtained quickly by exchanging the X value and the Z value in the coordinates of the original position center point, namely exchanging a 1 And c 1 The new equation of the first plane can be obtained quickly.
In this embodiment, considering that the parabola of the object moving in the air is mostly downward, that is, the opening is along the Y-axis direction, the projection onto the XOY plane or the ZOY plane is convenient for calculation.
S8: and judging whether the X coordinates of the two projection points in the three projection points are equal or not, if so, executing the step S9, and if not, executing the step S10.
S9: and fitting the three projection points through a circular equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point. Specifically, three projection points P 1 ’、P 2 ’、P 3 ' the coordinates are substituted into the circular equation (X-a 2 ) 2 +(Y-b 2 ) 2 ＝R 2 Obtaining a 2 、b 2 And R, the radius of curvature of the circle is the radius of the circle, so the radius of curvature r=r of the motion curve at the last projection point. Step S11 is performed.
S10: and fitting the three projection points by using a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point. Specifically, three projection points P 1 ’、P 2 ’、P 3 ' coordinates are substituted into the parabolic equation y=a 3 X 2 +b 3 X+c 3 Obtaining a 3 、b 3 And c 3 The method comprises the steps of carrying out a first treatment on the surface of the Then deriving an equation of the curvature radius r with respect to X according to the curvature formula to obtain r= [1+ (2 a) 3 X+b 3 ) 2 ] 3/2 /|2a 3 I (I); will P 3 Substituting the X coordinate value into the curvature radius equation can obtain the curvature radius of the motion curve at the last projection point. Step S11 is performed.
Further, not shown in the figure, if three projection points are substituted into the equation set of the parabolic equation without solution, step S9 is performed.
S11: and calculating the instantaneous speed of the target at the last projection point according to the curvature radius. Specifically, the centripetal force formula of curvilinear motion in physics is utilized: f=mv 2 Solving for object motion to point P 3 The instantaneous speed at' f=mgcos θ, since the centripetal force F is the component of the object in the curve normal direction (the present embodiment ignores the air resistance to the object motion, assuming that the object motion in air is only under gravity), and thus mv can be obtained 2 R=mgcos θ; the mass m can be reduced and then derived to give v 2 =rgcos θ, where v is the instantaneous speed, r is the radius of curvature calculated in step S9 or S10, g is the gravitational acceleration, θ is the motion curve at point P 3 The included angle between the normal direction and the gravity direction at the' position, the normal direction is perpendicular to the tangent line of the curve, and the included angle theta can be calculated by the existing calculation method.
S12: calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed; the frame interval time, i.e. the time interval between two adjacent frames, is numerically equal to the inverse of the frame rate in seconds. Specifically, the X-axis component v of the instantaneous velocity is calculated by orthogonal decomposition x The method comprises the steps of carrying out a first treatment on the surface of the Then according to the frame interval time andthe X-axis component of the instantaneous speed calculates the displacement of the next frame of the target in the X-axis direction, namely, let v x The displacement delta of the next frame of the target in the X-axis direction can be calculated by multiplying the frame interval time x 。
S13: calculating a plane predicted point P of the target on the motion curve according to the coordinate of the last projection point and the displacement in the X-axis direction 4 ' i.e. calculating the position of the object of the next frame on the motion curve, P 4 ' X-axis coordinate value, point P 3 The X-axis coordinate value plus the displacement in the X-axis direction is denoted as X 3 +Δ x Will x 3 +Δ x Substituting X into the parabolic equation in step S10, and taking the obtained Y as P 4 ' Y-axis coordinate value, i.e. P 4 The coordinates' in the XOY plane are (x 3 +Δ x ，Y Parabolic curve (x 3 +Δ x ))。
S14: and making a perpendicular line of the XOY plane through the plane prediction point, and calculating coordinates of an intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target. Specifically, a perpendicular to the XOY plane is drawn through the plane prediction point, and the perpendicular intersects the first plane Q at point P 4 The point P can be calculated by combining the equation of the first plane Q and the vertical line equation 4 Is defined by the coordinates of (a). Point P 4 I.e. the predicted position of the target in the next frame of RGB-D image.
Theoretically, the curve in the XOY plane is translated along the positive direction of the Z axis to form a curve plane, and the curve intersecting with the first plane Q is the motion curve of the target in the three-dimensional space. Because the direct calculation of the curve equation in the three-dimensional space is too complex, the curve equation is projected to the XOY plane first, and the motion curve on the XOY plane is calculated, so that the calculation process can be simplified.
The embodiment can predict the position of the object in the three-dimensional space, has high accuracy, does not need to model the motion trail of the target object in advance at one time, and models the position of any three frames in real time; the curve is utilized to fit the motion trail, then the curvature radius is utilized to calculate the instantaneous speed, and the error is small; meanwhile, the method can be used as an auxiliary means of object tracking tasks in a three-dimensional space for position verification.
Example two
The present embodiment is a computer-readable storage medium corresponding to the above embodiment, having stored thereon a computer program which, when executed by a processor, realizes the steps of:
acquiring an RGB image, and establishing a three-dimensional rectangular coordinate system according to the RGB image;
calculating coordinates of a position center point of the target in three frames of RGB-D images respectively;
determining a plane according to the three position center points to obtain a first plane, and calculating an equation of the first plane;
projecting the three position center points to an XOY plane to obtain three projection points, and calculating coordinates of the three projection points;
fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point;
calculating to obtain the instantaneous speed of the target at the last projection point according to the curvature radius;
calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed;
calculating a plane prediction point of the target on the motion curve according to the coordinate of the last projection point and the displacement in the X-axis direction;
and making a perpendicular line of the XOY plane through the plane prediction point, and calculating coordinates of an intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target.
Further, the establishing a three-dimensional rectangular coordinate system according to the RGB image specifically includes:
and establishing a three-dimensional rectangular coordinate system by taking the upper left corner of the RGB image as an origin, taking a ray passing through the origin and vertically downwards as a Y-axis positive direction, taking a ray passing through the origin and horizontally rightwards as an X-axis positive direction, and taking a ray passing through the origin and vertical to the inward direction of the RGB image as a Z-axis positive direction.
Further, after the three location center points are projected to the XOY plane to obtain three projection points and coordinates of the three projection points are calculated, the method further includes:
judging whether the three projection points are positioned on the same straight line or not;
if yes, transforming the projection surface;
and if not, executing the step of fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point.
Further, the fitting of the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the radius of curvature of the motion curve at the last projection point is specifically as follows:
judging whether the X coordinates of two projection points in the three projection points are equal in value or not;
if the motion curve exists, fitting the three projection points through a circular equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point;
if the motion curve does not exist, fitting the three projection points by using a parabolic equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point.
Further, according to the curvature radius, calculating to obtain the instantaneous speed of the target at the last projection point specifically includes:
calculating an included angle between the normal direction of the motion curve at the last projection point and the gravity direction;
calculating the instantaneous speed of the target at the last projection point according to a first formula, wherein the first formula is mv 2 R=mgcos θ, m is the target mass, v is the instantaneous velocity, r is the radius of curvature, g is the gravitational acceleration, θ is the included angle.
Further, the calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed specifically includes:
calculating an X-axis component of the instantaneous velocity by orthogonal decomposition;
and calculating the displacement of the target in the X-axis direction according to the frame interval time and the X-axis component.
Further, the perpendicular to the XOY plane is drawn through the plane prediction point, and coordinates of an intersection point of the perpendicular and the first plane are calculated, so that a predicted position point of the target is obtained specifically as follows:
making a perpendicular line of the XOY plane through the plane prediction point, and calculating an equation of the perpendicular line;
and calculating coordinates of an intersection point of the vertical line and the first plane according to the equation of the vertical line and the equation of the first plane to obtain a predicted position point of the target.
In summary, according to the target position prediction method and the storage medium based on the curvature radius provided by the invention, a curve in a three-dimensional space is fitted according to the position center point of the target object in the three-frame depth image, the curvature radius of the position of the point on the curve is calculated, then the instantaneous speed of the target in the third-frame depth image is calculated, and finally the position of the target object in the next-frame depth image is predicted by using the obtained instantaneous speed; the space curve is projected onto the plane by adopting a projection method, so that the calculation difficulty can be reduced, and the calculation efficiency can be improved; by utilizing curve fitting motion trail and utilizing curvature radius to calculate instantaneous speed, the calculated result error is ensured to be smaller, thereby ensuring the accuracy of subsequent position prediction. The method can predict the position of the object in the three-dimensional space, has high accuracy, does not need to model the motion track of the target object in advance at one time, and can be used as an auxiliary means for object tracking task in the three-dimensional space for position verification.
The foregoing description is only illustrative of the present invention and is not intended to limit the scope of the invention, and all equivalent changes made by the specification and drawings of the present invention, or direct or indirect application in the relevant art, are included in the scope of the present invention.
Claims (8)
1. A method for predicting a target location based on a radius of curvature, comprising:
acquiring an RGB image, and establishing a three-dimensional rectangular coordinate system according to the RGB image;
calculating coordinates of a position center point of the target in three frames of RGB-D images respectively;
determining a plane according to the three position center points to obtain a first plane, and calculating an equation of the first plane;
projecting the three position center points to an XOY plane to obtain three projection points, and calculating coordinates of the three projection points;
fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point;
calculating to obtain the instantaneous speed of the target at the last projection point according to the curvature radius;
calculating the displacement of the target in the X-axis direction according to the frame interval time and the instantaneous speed;
calculating a plane prediction point of the target on the motion curve according to the coordinate of the last projection point and the displacement in the X-axis direction;
and making a perpendicular line of the XOY plane through the plane prediction point, and calculating coordinates of an intersection point of the perpendicular line and the first plane to obtain a predicted position point of the target.
2. The method for predicting a target location based on a radius of curvature according to claim 1, wherein the creating a three-dimensional rectangular coordinate system from the RGB image is specifically:
and establishing a three-dimensional rectangular coordinate system by taking the upper left corner of the RGB image as an origin, taking a ray passing through the origin and vertically downwards as a Y-axis positive direction, taking a ray passing through the origin and horizontally rightwards as an X-axis positive direction, and taking a ray passing through the origin and vertical to the inward direction of the RGB image as a Z-axis positive direction.
3. The method according to claim 1, wherein after projecting the three position center points onto an XOY plane to obtain three projection points and calculating coordinates of the three projection points, further comprising:
judging whether the three projection points are positioned on the same straight line or not;
if yes, transforming the projection surface;
and if not, executing the step of fitting the three projection points by using a circular equation or a parabolic equation to obtain a motion curve, and calculating the curvature radius of the motion curve at the last projection point.
4. The method for predicting a target location based on a radius of curvature according to claim 1, wherein the fitting the three projection points with a circular equation or a parabolic equation to obtain a motion curve, and calculating the radius of curvature of the motion curve at the last projection point is specifically:
judging whether the X coordinates of two projection points in the three projection points are equal in value or not;
if the motion curve exists, fitting the three projection points through a circular equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point;
if the motion curve does not exist, fitting the three projection points by using a parabolic equation to obtain the motion curve, and calculating the curvature radius of the motion curve at the last projection point.
5. The method for predicting a target location based on a radius of curvature according to claim 1, wherein the calculating, according to the radius of curvature, an instantaneous speed of the target at a last projection point is specifically:
calculating an included angle between the normal direction of the motion curve at the last projection point and the gravity direction;
calculating the instantaneous speed of the target at the last projection point according to a first formula, wherein the first formula is mv 2 R=mgcos θ, m is the target mass, v is the instantaneous velocity, r is the radius of curvature, g is the gravitational acceleration, θ is the included angle.
6. The method according to claim 1, wherein the calculating the displacement of the target in the X-axis direction based on the frame interval time and the instantaneous speed is specifically:
calculating an X-axis component of the instantaneous velocity by orthogonal decomposition;
and calculating the displacement of the target in the X-axis direction according to the frame interval time and the X-axis component.
7. The method for predicting a target location based on a radius of curvature according to claim 1, wherein the making a perpendicular to an XOY plane by the plane predicting point and calculating coordinates of an intersection point of the perpendicular and the first plane, the obtaining a predicted location point of the target is specifically:
making a perpendicular line of the XOY plane through the plane prediction point, and calculating an equation of the perpendicular line;
and calculating coordinates of an intersection point of the vertical line and the first plane according to the equation of the vertical line and the equation of the first plane to obtain a predicted position point of the target.
8. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910514119.3A CN112085762B (en) | 2019-06-14 | 2019-06-14 | Target position prediction method based on curvature radius and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910514119.3A CN112085762B (en) | 2019-06-14 | 2019-06-14 | Target position prediction method based on curvature radius and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112085762A CN112085762A (en) | 2020-12-15 |
| CN112085762B true CN112085762B (en) | 2023-07-07 |
Family
ID=73733798
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910514119.3A Active CN112085762B (en) | 2019-06-14 | 2019-06-14 | Target position prediction method based on curvature radius and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112085762B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112577475A (en) * | 2021-01-14 | 2021-03-30 | 天津希格玛微电子技术有限公司 | Video ranging method capable of effectively reducing power consumption |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104835178A (en) * | 2015-02-02 | 2015-08-12 | 郑州轻工业学院 | Low SNR(Signal to Noise Ratio) motion small target tracking and identification method |
| WO2015172679A1 (en) * | 2014-05-14 | 2015-11-19 | 华为技术有限公司 | Image processing method and device |
| CN105701806A (en) * | 2016-01-11 | 2016-06-22 | 上海交通大学 | Depth image-based Parkinson tremor motion characteristic detection method and system |
| CN106875425A (en) * | 2017-01-22 | 2017-06-20 | 北京飞搜科技有限公司 | A kind of multi-target tracking system and implementation method based on deep learning |
| CN108986141A (en) * | 2018-07-03 | 2018-12-11 | 百度在线网络技术（北京）有限公司 | Object of which movement information processing method, device, augmented reality equipment and storage medium |
-
2019
- 2019-06-14 CN CN201910514119.3A patent/CN112085762B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015172679A1 (en) * | 2014-05-14 | 2015-11-19 | 华为技术有限公司 | Image processing method and device |
| CN104835178A (en) * | 2015-02-02 | 2015-08-12 | 郑州轻工业学院 | Low SNR(Signal to Noise Ratio) motion small target tracking and identification method |
| CN105701806A (en) * | 2016-01-11 | 2016-06-22 | 上海交通大学 | Depth image-based Parkinson tremor motion characteristic detection method and system |
| CN106875425A (en) * | 2017-01-22 | 2017-06-20 | 北京飞搜科技有限公司 | A kind of multi-target tracking system and implementation method based on deep learning |
| CN108986141A (en) * | 2018-07-03 | 2018-12-11 | 百度在线网络技术（北京）有限公司 | Object of which movement information processing method, device, augmented reality equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112085762A (en) | 2020-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110222581B (en) | Binocular camera-based quad-rotor unmanned aerial vehicle visual target tracking method | |
| CN109307508B (en) | Panoramic inertial navigation SLAM method based on multiple key frames | |
| CN113421289B (en) | High-precision vehicle track data extraction method for overcoming unmanned aerial vehicle shooting disturbance | |
| Gomez-Balderas et al. | Tracking a ground moving target with a quadrotor using switching control: nonlinear modeling and control | |
| CN109903346B (en) | Camera attitude detection method, device, device and storage medium | |
| CN111263960B (en) | Apparatus and method for updating high definition maps | |
| CN111476106B (en) | Method, system and device for real-time prediction of relative slope of straight road based on monocular camera | |
| CN109211241A (en) | The unmanned plane autonomic positioning method of view-based access control model SLAM | |
| CN107687850A (en) | A kind of unmanned vehicle position and orientation estimation method of view-based access control model and Inertial Measurement Unit | |
| WO2018133727A1 (en) | Method and apparatus for generating orthophoto map | |
| CN105606092B (en) | Indoor robot positioning method and system | |
| CN110455301A (en) | A Dynamic Scene SLAM Method Based on Inertial Measurement Unit | |
| CN110275179A (en) | A Map Construction Method Based on LiDAR and Vision Fusion | |
| Karakostas et al. | Shot type feasibility in autonomous UAV cinematography | |
| CN109542094B (en) | Visual Stabilization Control of Mobile Robots with Unexpected Images | |
| CN112085762B (en) | Target position prediction method based on curvature radius and storage medium | |
| CN116310799A (en) | A Dynamic Feature Point Removal Method Combining Semantic Information and Geometric Constraints | |
| CN116858269A (en) | Tobacco industry finished product warehouse flat warehouse inventory robot path optimization method based on laser SLAM | |
| CN110503684A (en) | Camera position and orientation estimation method and device | |
| Wang et al. | Pose and velocity estimation algorithm for UAV in visual landing | |
| CN108897345A (en) | A kind of method and system of control unmanned vehicle camera rotation | |
| CN118196205A (en) | On-line self-calibration method and system for external parameters of vehicle-mounted camera | |
| CN117782102A (en) | A fully automatic parking positioning and mapping method based on surround vision | |
| Cordes et al. | Accuracy evaluation of camera-based vehicle localization | |
| CN117308956A (en) | Mobile robot navigation control method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |