Target object intention prediction method and system

Abstract

The present invention provides a method for predicting the intention of a target object, which includes an information acquisition step, and a calculation and map data correspondence step. In the information acquisition step, the positioning information of the vehicle and multiple target object information of the target object are obtained, and the target object information corresponds to multiple time points on the time axis respectively, and each target object information includes the target object position and the target object speed. The calculation and map data correspondence step includes: corresponding the positioning information of the vehicle to the map data, corresponding the target object position corresponding to the last time point to the map data, and calculating the update speed of the target object with the target object speed in the target object information, corresponding the update speed to the map data, and predicting the future position of the target object on the map data at a future time point with the update speed. This helps to predict the intention of the target object.

Description

Technical Field

[0001] The present invention relates to a prediction method and a system thereof, and more particularly to a prediction method and a system thereof for analyzing the intention of a target object during driving. Background Art

[0002] Technological advances have enabled the development of vehicles to enter the automated stage of intelligent assisted driving or fully automated driving. Since the movement of vehicles is dynamic, it is necessary to avoid collisions with obstacles such as vehicles or pedestrians on the route. Therefore, some industry players have developed methods to predict the position or trajectory of target objects such as these obstacles to improve driving safety.

[0003] However, the urban environment is complex and cars and motorcycles are mixed, especially in the environment of intersections with multiple signal directions, where there are many obstacles and the trajectories are changeable, which easily leads to increased difficulty in analysis and judgment. In addition, the obstacle analysis on the market mostly predicts the situation at the next time point based on the situation at a single time point, which does not have time continuity and is also easy to lead to wrong judgment of obstacles.

[0004] In view of this, how to improve the analysis accuracy of obstacles and other target objects has become the goal of relevant industries. Summary of the invention

[0005] The present invention provides a method and system for predicting the intention of a target object. By using the vehicle positioning information and multiple target object information corresponding to multiple time points on a time axis, and mapping the vehicle positioning information and the target object information to a map, the method can help predict the intention of the target object and improve the analysis accuracy of the target object.

[0006] According to an embodiment of an aspect of the present invention, a method for predicting the intention of a target object is provided, which includes an information acquisition step (dataset obtaining step) and a calculation and map mapping step (calculating and map mapping step). In the information acquisition step, the vehicle positioning information of a vehicle traveling on a road and multiple target object information of at least one target object are obtained, and the target object information corresponds to multiple time points on a time axis during the travel, and each target object information includes a target object position and a target object speed. The calculation and map mapping step includes: mapping the vehicle positioning information to a map, mapping the target object position at the last time point to the map, and calculating an update speed of the at least one target object using the target object speed in the target object information, wherein the update speed is calculated by taking the target object speed within a time threshold, and the time threshold is defined as the target object speed at the last time point multiplied by an estimated value, and the estimated value is between 0 and 1, and the update speed is mapped to the map, and the update speed is used to predict the future position of the at least one target object on the map at a future time point.

[0007] By mapping the vehicle and the target to the map data, the position of the target can be clearly known, which helps to judge the target's intention. In addition, by calculating the target's speed at different time points to obtain the target's update speed, the target's future position can be predicted, which can make the prediction of the target have temporal continuity and improve the accuracy of target analysis.

[0008] According to the aforementioned target object intention prediction method, a future trajectory selection step may be further included, wherein the vehicle may include multiple future trajectories located on the map data, and one of the future trajectories is selected according to the future position of the aforementioned at least one target object to avoid the aforementioned at least one target object.

[0009] According to the aforementioned target object intention prediction method, in the calculation and map data correspondence steps, a region of interest on the map data can be adjusted according to the future trajectory, and when the aforementioned at least one target object is located within the region of interest, the update speed of the aforementioned at least one target object is calculated.

[0010] According to the aforementioned target object intention prediction method, the future position of a vehicle at the last future time point in the future trajectory can be found, and the future position of the vehicle plus a first set distance is used as a front boundary of the area of ​​interest, and the rear end of a vehicle body of the vehicle plus a second set distance is used as a rear boundary of the area of ​​interest, the left side of the vehicle body plus the width of a lane is used as a left boundary of the area of ​​interest, and the right side of the vehicle body plus the width of a lane is used as a right boundary of the area of ​​interest.

[0011] According to the aforementioned target object intention prediction method, the aforementioned at least one target object can be classified into a vehicle category, a pedestrian category or a pedestrian category, and different estimation values ​​are set for the vehicle category, the pedestrian category and the pedestrian category.

[0012] According to the aforementioned target object intention prediction method, the targets may be classified according to the position of the at least one target object on the image data, the length and width of the at least one target object, or the length-to-width ratio of the at least one target object.

[0013] According to the aforementioned target object intention prediction method, the future time point can be set equal to the time threshold.

[0014] According to the aforementioned target object intention prediction method, in the calculation and image data corresponding steps, a weighted moving average method can be used to calculate the update speed.

[0015] According to the aforementioned target object intention prediction method, in the calculation and image data corresponding steps, a real-time dynamic positioning technology can be used to obtain the vehicle positioning information.

[0016] According to another embodiment of the present invention, a target intention prediction system is provided, which is applied to the target intention prediction method, and the target intention prediction system includes a vehicle, at least one sensor and a processor. The at least one sensor is disposed on the vehicle and is used to detect the at least one target, and the processor signal is connected to the at least one sensor to obtain target information.

[0017] According to the aforementioned target intention prediction system, wherein the at least one sensor may have an optical sensor structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A block diagram illustrating a method for predicting target intention according to an embodiment of the present invention is shown;

[0019] Figure 2 Draw Figure 1 A flowchart of the steps of the target object intention prediction method;

[0020] Figure 3 Draw Figure 1 A schematic diagram of a method for predicting target object intention corresponding to a graph; and

[0021] Figure 4 FIG. 4 is a schematic diagram showing the architecture of a target object intention prediction system according to another embodiment of the present invention.

[0022]

Explanation of symbols

[0023] 100 Target Intention Prediction Method

[0024] Steps to obtain 110 information

[0025] 120 calculation and corresponding steps of graphs

[0026] 130 Future Trajectory Selection Steps

[0027] K1 Maps

[0028] H1 car

[0029] R1 road

[0030] Steps S01, S02, S03

[0031] Steps S04, S05, and S06

[0032] Steps S07, S08, S09

[0033] Steps S10, S11, S12

[0034] G1, G2, G3 targets

[0035] G4, G5 target

[0036] The future trajectory of D1, D2 and D3

[0037] W1 Area of ​​interest

[0038] P1 Target position

[0039] P2 The future position of the target

[0040] 200 Target Intention Prediction System

[0041] 210 car

[0042] 220 Sensors

[0043] 230 processors DETAILED DESCRIPTION

[0044] Embodiments of the present invention will be described below with reference to the accompanying drawings. For the sake of clarity, many practical details will be described together in the following description. However, the reader should understand that these practical details should not be used to limit the present invention. That is, in some embodiments of the present invention, these practical details are not necessary. In addition, in order to simplify the drawings, some conventional structures and elements will be depicted in a simple schematic manner in the drawings; and repeated elements may be represented by the same number or similar number.

[0045] In addition, in this article, when a certain element (or mechanism or module, etc.) is "connected", "set" or "coupled" to another element, it may refer to that the element is directly connected, directly set or directly coupled to another element, and it may also refer to that a certain element is indirectly connected, indirectly set or indirectly coupled to another element, that is, there are other elements between the element and the other element. When it is clearly stated that a certain element is "directly connected", "directly set" or "directly coupled" to another element, it means that there are no other elements between the element and the other element. The terms first, second, third, etc. are only used to describe different elements or components, and there is no restriction on the elements / components themselves. Therefore, the first element / component can also be renamed as the second element / component. Moreover, the combination of elements / components / mechanisms / modules in this article is not a generally known, conventional or conventional combination in this field. Whether the elements / components / mechanisms / modules themselves are conventional cannot be used to determine whether their combination relationship is easy to be easily completed by ordinary knowledge in the technical field.

[0046] See also Figure 1 ,in Figure 1 FIG. 1 is a block diagram of a method 100 for predicting target intention according to an embodiment of the present invention. The method 100 for predicting target intention includes an information acquisition step 110 and a calculation and image data correspondence step 120 .

[0047] In the information acquisition step 110 , a vehicle positioning information of a vehicle and a plurality of target object information of a target object are acquired. The target object information respectively corresponds to a plurality of time points on a time axis, and each target object information includes a target object position and a target object speed.

[0048] The calculation and map data correspondence step 120 includes: corresponding the vehicle positioning information to a map data, corresponding the target object position at the last time point to the map data, and calculating an update speed of the target object using the target object speed in the target object information, corresponding the update speed to the map data, and using the update speed to predict a future position of the target object on the map data at at least one future time point.

[0049] Thus, by mapping the vehicle and the target object to the map data, the position of the target object can be clearly known, which helps to judge the target object's intention. In addition, by calculating the target object's speed at different time points to obtain the target object's update speed, the target object's future position can be predicted, so that the prediction of the target object has temporal continuity, which can improve the accuracy of the target object analysis. The details of the target object intention prediction method 100 will be described in more detail later.

[0050] The vehicle is traveling on a road, and the vehicle positioning information may be obtained by a real-time kinematic positioning technology (Real Time Kinematic; RTK), so the vehicle positioning information may include a longitude, a latitude, and a heading angle of the vehicle.

[0051] The target object information may be obtained by at least one sensor installed on the vehicle. In other embodiments, the target object information may also be obtained from environmental information transmitted by vehicle to everything (V2X), but is not limited thereto.

[0052] Therefore, in the information acquisition step 110, when the vehicle is moving, the sensor will continuously detect the target object and obtain the target object information corresponding to the target object at each time point, and then store the target object information in a matrix form in a processor installed on the vehicle. It should be particularly noted here that the target object position included in each target object information obtained by the sensor refers to its relative position with the vehicle, and the target object speed refers to its relative speed with the vehicle.

[0053] In the calculation and map data correspondence step 120, the main purpose is to correspond all the information to the map data. More specifically, the map data has rich map information, and preferably has high-precision map information, such as road markings, number of lanes, sidewalks and buildings. Therefore, when all the information corresponds to the map data, the location of the vehicle and the target object on the map data can be clearly understood, which can help to improve the accuracy of intention prediction. For example, when the target object is located on the sidewalk, its intention prediction should be to move along the sidewalk; when the target object is located at the intersection, it is inferred that its intention is to cross the road. In addition, if the target object is on the sidewalk, it can be determined to be a pedestrian or a quasi-pedestrian, so the speed change of the target object is lower, but the maneuverability of direction change is higher. In this way, through the map data, intention prediction can be effectively assisted.

[0054] In one embodiment, the map data may be stored in advance in the processor of the vehicle, or obtained in real time through the Internet of Vehicles, but is not limited thereto. The vehicle positioning information obtained through real-time dynamic positioning technology may include longitude, latitude and heading angle. The map data may be rotated by the heading angle and then translated by the longitude and latitude, so that the vehicle positioning information and the map data can correspond.

[0055] Since the target position contained in each target information refers to its relative position with respect to the vehicle, a coordinate transfer formula is required, as shown in formula (1), to transfer the target position to the map data, and the origin after the transfer is the position of the vehicle.

[0056]

[0057] Where x and y represent the original target position, x' and y' represent the target position transferred to the image data (corresponding to the X-axis and Y-axis), θ is the heading angle, and the X-axis is 0 degrees.

[0058] In the calculation and map data correspondence step 120, the target object position at the last time point can be first matched to the map data, and then the update speed of the target object is calculated, and finally the update speed is matched to the map data; or the update speed of the target object is first calculated, and then the target object position and update speed at the last time point are simultaneously matched to the map data, but the present invention is not limited to this.

[0059] Preferably, the update speed may be calculated using a weighted moving average (WMA) method, and the weighted moving average method is shown in formula (2).

[0060]

[0061] Among them Vwma x Represents the component of the update speed on the X-axis, Vwma y Represents the component of the update speed on the Y axis, i is a positive integer, n is the number of time points, V xi Represents the component of the target velocity on the X-axis at each time point, V yi represents the component of the target speed corresponding to each time point on the Y axis. For example, the target speeds from the 1st time point to the 10th time point can be substituted into formula (2) for calculation, so that n=10. When the target speed at the 11th time point is obtained, n=10 is still maintained, the target speed data at the 1st time point is deleted, and the target speed at the 2nd time point is changed to the target speed at the 1st time point. And so on, the target speed at the 11th time point is changed to the target speed at the 10th time point. In other words, the oldest target speed is deleted and the latest target speed is added to maintain the same n.

[0062] After calculating the update speed, the update speed can be substituted into equation (3) to correspond to the image data.

[0063]

[0064] Among them Vwma x 'Represents the update speed transferred to the graphics data on the X-axis, Vwma y 'Represents the component on the Y-axis of the update speed transferred to the graphics data.

[0065] Thus, by multiplying the update speed transferred to the map data by the future time point, for example, 1 second, the future position of the target object on the map data after 1 second can be obtained.

[0066] In the calculation and image data correspondence step 120, the target speed within a time threshold can be used to calculate the update speed, and the time threshold is defined as the target speed at the last time point multiplied by an estimated value, and the estimated value is between 0 and 1. Through the setting of the time threshold, the target can adjust the time threshold according to the target speed at the last time point (i.e., the current speed), so that the update speed is more in line with the actual situation, thereby improving the accuracy of the target intention prediction method 100.

[0067] In one embodiment, the target object intention prediction method 100 can classify the target object into a pedestrian category, a pedestrian category, or a vehicle category in the calculation and image data corresponding step 120, and set different estimation values ​​for the vehicle category, the pedestrian category, and the pedestrian category. If the target object is classified as a vehicle category, the estimation value is set to 1 / 5 second; if the target object is classified as a pedestrian category, the estimation value is set to 1 / 8 second; if the target object is classified as a pedestrian category, the estimation value is set to 1 / 10 second.

[0068] When classifying the target object, the target object may be classified according to the position of the target object on the map, the length and width of the target object, or the length-to-width ratio of the target object. In one embodiment, if the target object is located on the sidewalk, the target object is determined to be a pedestrian. If the target object is not located on the sidewalk, it is further determined whether the length and width of the target object are greater than 4 meters and 1.5 meters respectively. If so, the target object is determined to be a vehicle. Otherwise, it is further determined whether the length-to-width ratio of the target object is greater than 2:1. If so, the target object is determined to be a pedestrian-like category, such as a motorcycle or a bicycle. Otherwise, the target object is determined to be a pedestrian. The length and width of the target object can be detected by the vehicle with lidar, and then converted to the map using equations (4) and (5) to classify the target object.

[0069] H r =W o ×cosθ 2 +H o ×sinθ 2 (4).

[0070] W r =W o ×sinθ 2 +H o ×cosθ 2 (5).

[0071] Among them, H r Indicates the converted length, W r Indicates the width after conversion, H o Indicates the original length, W o Indicates the original width.

[0072] After classifying the target, different estimated values ​​can be set for different target categories, taking into account factors such as the flexibility, degree of freedom, and whether the path is easy to change. Therefore, when the target is classified as a vehicle and the target speed at the last time point is 40 km / h (11.1 m / s), the estimated value is 1 / 5 second, so the time threshold can be set to 2.2 seconds, and the speed of all target objects at all time points within 2.2 seconds from the last time point is substituted into formula (2) for calculation. When the target is classified as a pedestrian and the target speed at the last time point is 40 km / h (11.1 m / s), the estimated value is 1 / 8 second, so the time threshold can be set to 1.3 seconds, and the speed of all target objects at all time points within 1.3 seconds from the last time point is substituted into formula (2) for calculation. When the target object is classified as a pedestrian and the target speed at the last time point is 16 km / h (4.4 m / s), the estimated value is 1 / 10 second, so the time threshold can be set to 0.44 seconds, and the speeds of all target objects at all time points within 0.44 seconds from the last time point are substituted into formula (2) for calculation, but the present invention is not limited to this.

[0073] In addition, the future time point can be set equal to the time threshold. Therefore, when the target object is classified as a vehicle, the future time point can be set to 2.2 seconds, that is, the coordinate of the future position on the X-axis is equal to x'+(Vwma x ')x2.2, the coordinate of the future position on the Y axis is equal to y'+(Vwma y ')x2.2. Similarly, when the target object is classified as a pedestrian, the future time point can be set to 1.3 seconds, that is, the coordinate of the future position on the X-axis is equal to x'+(Vwma x ')x1.3, the coordinate of the future position on the Y axis is equal to y'+(Vwma y ')x1.3; When the target object is classified as a pedestrian, the future time point can be set to 0.44 seconds, that is, the coordinate of the future position on the X-axis is equal to x'+(Vwma x ')x0.44, the coordinate of the future position on the Y axis is equal to y'+(Vwma y ')x0.44.

[0074] In addition, Figure 1 In the embodiment of the present invention, the target object intention prediction method 100 may further include a future trajectory selection step 130. The vehicle includes multiple future trajectories located on the map data, and one of the future trajectories may be selected according to the future position of the target object to avoid the target object.

[0075] Through the maximum and minimum values ​​of the vehicle's speed, heading angle, and path curvature, multiple future trajectories that the vehicle can travel can be obtained, that is, each trajectory contains the future position of the vehicle corresponding to multiple future time points. Since the future position of the target object at the future time point has been predicted, a suitable future trajectory can be selected, and the future position of the vehicle can be different from the future position of the target object at the same future time point to avoid the target object.

[0076] In the calculation and map data correspondence step 120, a region of interest (ROI) on the map data can be adjusted according to the future trajectory. When the target object is within the region of interest, the update speed of the target object is calculated. In the case of too many targets, if the targets are analyzed or calculated one by one, the efficiency of the processor in the vehicle is likely to be reduced. Therefore, the area within a certain distance from the vehicle can be set as the region of interest. The targets within the region of interest have the possibility of colliding with the vehicle, while the targets outside the region of interest have no possibility of colliding with the vehicle. Therefore, by only analyzing the targets within the region of interest, the efficiency of the processor can be improved.

[0077] The scope of the region of interest may include the range of multiple future trajectories. For example, the future position of a vehicle at the last future time point in the future trajectory (the farthest one) may be found, and the future position of the vehicle mentioned above plus a first set distance is used as a front boundary of the region of interest. The first set distance may be, for example, 3 meters, and the rear end of a body of the vehicle plus a second set distance is used as a rear boundary of the region of interest. The second set distance may be, for example, 3 meters; the left side of the vehicle plus the width of a lane is used as a left boundary of the region of interest, and the right side of the vehicle plus the width of a lane is used as a right boundary of the region of interest, including but not limited to this. In this way, the scope of the region of interest can be adjusted through the dynamics of the vehicle and the future trajectory, with higher flexibility.

[0078] See also Figures 2 to 3 , and with reference Figure 1 ,in Figure 2 Draw Figure 1 A flowchart of the steps of the target object intention prediction method 100, Figure 3 Draw Figure 1 The target object intention prediction method 100 corresponds to a schematic diagram of the graph K1.

[0079] The vehicle H1 is traveling on a road R1. The vehicle H1 can continuously perform dynamic positioning and can proceed to step S02 to obtain the vehicle positioning information. At the same time, step S03 is performed to detect multiple targets G1, G2, G3, G4, and G5. In combination with step S01, the map data K1 is obtained in real time. Step S04 can be performed to store the vehicle positioning information and the target positions of the targets G1, G2, G3, G4, and G5 ( Figure 3 The target position P1 of the target G1 is only indicated in the figure and corresponds to the map K1.

[0080] The target information detected in step S03 can be stored and processed for subsequent use. In one embodiment, multiple targets G1, G2, G3, G4, and G5 are numbered from near to far, that is, target G1 is closest to the vehicle H1, and target G5 is farthest from the vehicle H1. If the same target (such as target G1) is detected at multiple consecutive time points, the cumulative number of target G1 will be accumulated. On the contrary, if the target (such as target G2) disappears, the cumulative number of target G2 will be reset to zero and recalculated, as shown in Table 1.

[0081] Table 1. Cumulative number of targets at different time points

[0082] G1 cumulative times G2 cumulative times G3 cumulative times G4 cumulative number of times G5 cumulative times Time point 1 1 1 1 1 - Time point 2 2 2 2 - - Time point 3 3 - 3 1 - Time point 4 4 1 - - 1

[0083] Specifically, as shown in Table 1, when the targets G1, G2, G3, and G4 are detected at time point 1, the target information of the targets G1, G2, G3, and G4 is stored in the processor, and the cumulative counts corresponding to the targets G1, G2, G3, and G4 are all 1. When the targets G1, G2, and G3 are detected at time point 2, the target information of the targets G1, G2, and G3 is stored in the processor, and all target information corresponding to the undetected target G4 before time point 2 is cleared. When the targets G1, G3, and G4 are detected at time point 3, the target information of the targets G1, G3, and G4 is stored in the processor, and all target information corresponding to the undetected target G2 before time point 3 is cleared; and since the target G4 disappeared at time point 2, the cumulative count will be reset to zero and recalculated, so the cumulative count of the target G4 corresponding to time point 3 is 1. It should be noted that the target G4 at time point 1 and the target G4 at time point 3 may be different objects, but are given the same number only because of the distance. In addition, when the targets G1, G2, and G5 are detected at time point 4, the target information of the targets G1, G2, and G5 is stored in the processor, and all the target information of the undetected targets G3 and G4 before time point 4 is cleared; and because the target G2 disappeared at time point 3, the cumulative number will be reset to zero and recalculated, so the cumulative number of the target G2 corresponding to time point 4 is 1. It should be noted that the above data processing can be completed before step S09, and it does not necessarily have to be completed in step S03.

[0084] Step S05 can be executed in advance, and it can obtain multiple future trajectories D1, D2, and D3 that the vehicle H1 can travel based on the maximum and minimum values ​​of the vehicle speed, the heading angle, and the path curvature. Therefore, when step S04 is completed, the information of step S05 is obtained, and step S06 can be performed to adjust the area of ​​interest W1 based on the future trajectories D1, D2, and D3.

[0085] Then, step S07 is performed to confirm whether the target objects G1, G2, G3, G4, G5 are located in the region of interest W1. Figure 3 As shown, after only the target object G1 is located in the region of interest W1, step S08 may be performed to classify the target object G1.

[0086] Through the above classification method, the target object G1 can be classified as a pedestrian. According to the above setting of the time threshold, the time threshold can be set to 0.44 seconds.

[0087] Therefore, assuming that the vehicle H1 is detected once every 0.02 seconds, if the cumulative number of target object G1 after data processing is equal to 22, it means that it has been continuously detected within the time threshold of 0.44 seconds. Therefore, the target object speed of these 22 times is substituted into formula (2) for calculation, and n=22 is set to perform step S09 to obtain the updated speed.

[0088] Next, step S10 is performed to convert the update speed using equation (3), and then step S11 is performed to predict the future position P2 of the target object at a future time point of 0.44 seconds based on the update speed of the map data K1 and the target object position P1 at the last time point.

[0089] But can Figure 3 As shown, by the future time and the future position P2 of the target object, step S12 is performed to select one of the three future trajectories D1, D2, and D3. It should be particularly noted that the future trajectories D2 and D3 do not intersect with the target object G1, but the future trajectory D2 does not need to change the speed or direction, so it is the best choice and can be selected first.

[0090] See also Figure 4 ,in Figure 4 FIG. 2 is a schematic diagram of the structure of a target object intention prediction system 200 according to another embodiment of the present invention. Figure 1 The target intention prediction method 100 is provided. The target intention prediction system 200 comprises a vehicle 210, at least one sensor 220 and a processor 230. The at least one sensor 220 is disposed on the vehicle 210 and used to detect the target. The processor 230 is signal-connected to the at least one sensor 220 to obtain target information.

[0091] exist Figure 4 In the embodiment, the number of the sensors 220 may be two, and preferably, the sensor 220 has an optical sensor structure, such as an optical radar structure, but in other embodiments, it is not limited thereto.

[0092] It can be seen from the above embodiments that the present invention has the following advantages.

[0093] 1. By mapping the vehicle and the target to the map data, the future change capability of static or dynamic targets around the vehicle can be effectively predicted, and the movement prediction of the target in a complex environment (such as an urban area) can be solved.

[0094] Second, through the future trajectory selection step, the best future trajectory can be selected to avoid the target object.

[0095] Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Anyone familiar with the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the scope defined in the attached claims.

Claims

1. A method for predicting target intention, characterized in that: Include: an information acquisition step, acquiring a vehicle positioning information of a vehicle traveling on a road and a plurality of target object information of at least one target object, the target object information respectively corresponding to a plurality of time points on a time axis during traveling, and each of the target object information includes a target object position and a target object speed; and A calculation and image data corresponding step includes: Corresponding the vehicle positioning information to a map; Mapping the target object position at the last time point to the map data; and An update speed of the at least one target is calculated using the target speeds in the target information, wherein the update speed is calculated by taking the target speeds within a time threshold, and the time threshold is defined as the target speed at the last time point multiplied by an estimated value, and the estimated value is between 0 and 1. The update speed is mapped to the map data, and the update speed is used to predict a future position of the at least one target on the map data at a future time point.

2. The target object intention prediction method according to claim 1, characterized in that: Also includes: A future trajectory selection step, wherein the vehicle includes a plurality of future trajectories located on the map data, and one of the future trajectories is selected according to the future position of the at least one target object to avoid the at least one target object.

3. The target object intention prediction method according to claim 2, characterized in that: In the calculation and map data correspondence step, a region of interest on the map data is adjusted according to the future trajectories, and when the at least one target object is located within the region of interest, the update speed of the at least one target object is calculated.

4. The target object intention prediction method according to claim 3, characterized in that: Find the future position of the vehicle at the last future time point in the future trajectory, and use the future position of the vehicle plus a first set distance to form a front boundary of the area of ​​interest, and use the rear end of a body of the vehicle plus a second set distance to form a rear boundary of the area of ​​interest, use the left side of the vehicle body plus a lane width to form a left boundary of the area of ​​interest, and use the right side of the vehicle body plus a lane width to form a right boundary of the area of ​​interest.

5. The target object intention prediction method according to claim 1, characterized in that: The at least one target object is classified into a vehicle type, a pedestrian type or a pedestrian type, and different estimation values ​​are set for the vehicle type, the pedestrian type and the pedestrian type.

6. The target object intention prediction method according to claim 5, characterized in that: The target object is classified according to the position of the at least one target object on the image data, the length and width of the at least one target object, or the length-to-width ratio of the at least one target object.

7. The target object intention prediction method according to claim 6, characterized in that: The future time point is set equal to the time threshold.

8. The target object intention prediction method according to claim 1, characterized in that: In the calculation and data correspondence step, the update speed is calculated using a weighted moving average method.

9. The method for predicting target intention according to claim 1, characterized in that: In the calculation and map data corresponding step, the vehicle positioning information is obtained by a real-time dynamic positioning technology.

10. A target object intention prediction system, characterized in that: Applied to the target object intention prediction method as claimed in claim 1, the target object intention prediction system comprises: the vehicle; At least one sensor, disposed on the vehicle and used to detect the at least one target object; and A processor is signal-connected to the at least one sensor to obtain the target information.

11. The target object intention prediction system according to claim 10, characterized in that: The at least one sensor has an optical sensor structure.

Images