CN114312829A - Pedestrian trajectory prediction method and device, electronic equipment and storage medium - Google Patents
Pedestrian trajectory prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114312829A CN114312829A CN202111475997.2A CN202111475997A CN114312829A CN 114312829 A CN114312829 A CN 114312829A CN 202111475997 A CN202111475997 A CN 202111475997A CN 114312829 A CN114312829 A CN 114312829A
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- model
- track
- special
- normal
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000004590 computer program Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012417 linear regression Methods 0.000 claims description 5
- 238000003062 neural network model Methods 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 claims description 3
- 230000003993 interaction Effects 0.000 abstract description 9
- 230000007774 longterm Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a pedestrian trajectory prediction method and device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring an actual track of a target pedestrian, a historical predicted track obtained based on a normal pedestrian model and a historical predicted track obtained based on a special pedestrian model; obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the normal pedestrian model; obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model; and generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model. The method and the device can accurately predict the tracks of normal pedestrians and special pedestrians, improve the precision of the automatic driving interaction strategy, and reduce the probability of dangerous accidents of colliding with the special pedestrians.
Description
Technical Field
The present disclosure relates to the field of automatic driving technologies, and in particular, to a method and an apparatus for predicting a pedestrian trajectory, an electronic device, and a storage medium.
Background
At present, the algorithm based on big data can predict the pedestrian track (for example, the walking position of the target pedestrian n seconds in the future) or the intention (for example, the target pedestrian does not need to cross the road), and can better give the prediction result according with the natural distribution and the interaction rule according with the natural distribution (for example, the person tends to stop at the roadside or the like, and the vehicle passes through the road, but does not simply move in a straight line at a constant speed). Compared with the traditional artificial intelligence method based on pattern recognition or rules, the method provides more accurate income or loss experience estimation for the automatic driving interaction strategy.
However, for special pedestrians (e.g., blind people, children playing on the road, or people watching a mobile phone to cross the road), since the proportion of the population number of such special pedestrians to the total data amount is very small, when the trajectory or the intended probability distribution is fitted by the big data algorithm, the overall distribution of the data (i.e., how most normal pedestrians will do) is inclined, and the individual distribution of the special pedestrians (e.g., how special pedestrians will do) is ignored, thereby resulting in serious under-estimation.
Disclosure of Invention
Therefore, the technical problem solved by the embodiments of the present application is to provide a pedestrian trajectory prediction method and apparatus, an electronic device, and a storage medium, which can accurately predict trajectories of normal pedestrians and special pedestrians, further improve the accuracy of an automatic driving interaction strategy, and effectively reduce the probability of dangerous accidents of colliding with special pedestrians.
In order to solve the technical problem, the technical scheme adopted by the application specifically comprises the following steps:
in a first aspect, an embodiment of the present application provides a method for predicting a pedestrian trajectory, including:
acquiring an actual track of a target pedestrian, a historical predicted track obtained based on a normal pedestrian model and a historical predicted track obtained based on a special pedestrian model;
obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the normal pedestrian model;
obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model;
and generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
Further, the normal pedestrian model is a deep neural network model, and the special pedestrian-based model is a rule logic model.
Further, the normal pedestrian model is a linear regression model, and the special pedestrian-based model is a rule logic model.
Further, the normal pedestrian model is an SVM model, and the special pedestrian-based model is a rule logic model.
Preferably, obtaining the similarity of the trajectories based on the normal pedestrian model according to the actual trajectory of the target pedestrian and the historical predicted trajectory obtained based on the normal pedestrian model includes:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a normal pedestrian model to the prediction positions D1, D2, …, Di, … and Dk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Si ^ e (-a × Di), wherein Di is the distance difference between the ith adoption time point Ti and the predicted position Di of the current time T0 and the current position G0 of the current time T0, and a is a prediction correction parameter;
according to the SUM { calcsore (Di, G0) × e (-Ti) | i ═ 1,2, …, k }, the trajectory similarity S based on the normal pedestrian model is obtained.
More preferably, the obtaining of the similarity of the trajectory based on the special pedestrian model according to the actual trajectory of the target pedestrian and the historical predicted trajectory based on the special pedestrian model includes:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a special pedestrian model to the prediction positions R1, R2, …, Ri, … and Rk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Ri ═ e (-b × Ri), wherein Ri is the distance difference between the ith adoption time point Ti and the predicted position Ri of the current time T0 and the current position G0 of the current time T0, and b is a predicted correction parameter;
according to the formula, the trajectory similarity R based on the special pedestrian model is obtained according to the formula of R { calcsore (Ri, G0) × e (-Ti) | i ═ 1,2, …, k }.
More preferably, generating the predicted trajectory of the target pedestrian based on the normal pedestrian model or the special pedestrian model according to the comparison result of the similarity of the trajectory based on the normal pedestrian model and the similarity of the trajectory based on the special pedestrian model includes:
acquiring the ratio of the track similarity S based on the normal pedestrian model to the track similarity R based on the special pedestrian model according to the track similarity S based on the normal pedestrian model and the track similarity R based on the special pedestrian model;
judging whether the ratio is smaller than a preset threshold value:
if yes, generating a target pedestrian predicted track based on the special pedestrian model;
and if not, generating a target pedestrian predicted track based on the normal pedestrian model.
Further, before obtaining the actual track of the target pedestrian, the historical predicted track obtained based on the normal pedestrian model and the historical predicted track obtained based on the special pedestrian model, the method further comprises the following steps:
acquiring environmental information in a preset area;
extracting target pedestrian information in the environment information;
and acquiring the actual track of the target pedestrian according to the target pedestrian information.
In a second aspect, an embodiment of the present application provides a pedestrian trajectory prediction device, including:
the track acquisition module is used for acquiring the actual track of the target pedestrian, the historical prediction track obtained based on the normal pedestrian model and the historical prediction track obtained based on the special pedestrian model;
the first processing module is used for obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the normal pedestrian model;
the second processing module is used for obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the special pedestrian model;
and the third processing module is used for generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
In a third aspect, an embodiment of the present application provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements the steps of the pedestrian trajectory prediction method according to any one of the above items when executing the computer program.
In a fourth aspect, an embodiment of the present application provides a storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the steps of the pedestrian trajectory prediction method described in any one of the above.
In summary, compared with the prior art, the beneficial effects brought by the technical scheme provided by the embodiment of the present application at least include:
1. according to the embodiment of the application, the track similarity based on the normal pedestrian model is obtained through the actual track of the target pedestrian and the historical prediction track obtained based on the normal pedestrian model; obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model; the method comprises the steps of generating a target pedestrian predicted track based on a normal pedestrian model or a special pedestrian model according to a comparison result of track similarity based on the normal pedestrian model and track similarity based on the special pedestrian model, judging the similarity between an actual track of a target pedestrian and a historical predicted track obtained based on the normal pedestrian model and the historical predicted track obtained based on the special pedestrian model respectively, and judging whether the target pedestrian belongs to a normal pedestrian or a special pedestrian according to the comparison result of the similarity, so that the target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model is generated to be used as a basis for an automatic driving interaction strategy by an automatic driving vehicle, the accuracy of the automatic driving interaction strategy is corrected, and the occurrence probability of dangerous accidents is effectively reduced.
2. According to the embodiment of the application, the environment information in the preset area is obtained, the target pedestrian information in the environment information is extracted, the actual track of the target pedestrian is obtained according to the target pedestrian information, unnecessary noise information can be removed from the environment information, and the actual track of the target pedestrian is identified, so that the accuracy of the pedestrian track prediction method is further improved.
Drawings
Fig. 1 is a flowchart illustrating a method for predicting a pedestrian trajectory according to a first exemplary embodiment of the present application.
Fig. 2 is a schematic structural diagram of a pedestrian trajectory prediction apparatus according to a ninth exemplary embodiment of the present application.
Fig. 3 is a schematic structural diagram of an electronic device according to a tenth exemplary embodiment of the present application.
Detailed Description
The present embodiment is only for explaining the present application, and it is not limited to the present application, and those skilled in the art can make modifications of the present embodiment without inventive contribution as needed after reading the present specification, but all of them are protected by patent law within the scope of the claims of the present application.
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The terms "comprises," "comprising," or any other variation thereof, in the description and claims of this application, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
In the embodiments of the present application, words such as "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "e.g.," is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the word "exemplary" or "such as" is intended to present concepts related in a concrete fashion.
The terms "first," "second," and the like in this application are used for distinguishing between similar items and items that have substantially the same function or similar functionality, and it should be understood that "first," "second," and "nth" do not have any logical or temporal dependency or limitation on the number or order of execution.
The embodiments of the present application will be described in further detail with reference to the drawings attached hereto.
Fig. 1 is a pedestrian trajectory prediction method according to a first exemplary embodiment of the present application, including:
acquiring an actual track of a target pedestrian, a historical predicted track obtained based on a normal pedestrian model and a historical predicted track obtained based on a special pedestrian model;
obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the normal pedestrian model;
obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model;
and generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
In the first exemplary embodiment of the present application, the trajectory similarity based on the normal pedestrian model is obtained by predicting the trajectory based on the actual trajectory of the target pedestrian and the history obtained based on the normal pedestrian model; obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model; the method comprises the steps of generating a target pedestrian predicted track based on a normal pedestrian model or a special pedestrian model according to a comparison result of track similarity based on the normal pedestrian model and track similarity based on the special pedestrian model, judging the similarity between an actual track of a target pedestrian and a historical predicted track obtained based on the normal pedestrian model and the historical predicted track obtained based on the special pedestrian model respectively, and judging whether the target pedestrian belongs to a normal pedestrian or a special pedestrian according to the comparison result of the similarity, so that the target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model is generated to be used as a basis for an automatic driving interaction strategy by an automatic driving vehicle, the accuracy of the automatic driving interaction strategy is corrected, and the occurrence probability of dangerous accidents is effectively reduced.
Specifically, the normal pedestrian model is a deep neural network model, and the special pedestrian-based model is a rule logic model.
In other embodiments, the normal pedestrian model is a linear regression model and the special pedestrian-based model is a regular logic model.
In other embodiments, the normal pedestrian model is an SVM model, and the special pedestrian-based model is a regular logic model.
The predicted trajectory obtained by using the deep neural network model has the highest precision which can reach 70 percent compared with a linear regression model and an SVM model; secondly, the accuracy of the linear regression model can reach 55%, and finally, the accuracy of the SVM model can reach 40%.
A second exemplary embodiment of the present application provides a method for predicting a pedestrian trajectory, which is a further improved solution based on the first exemplary embodiment of the present application shown in fig. 1, and is different from the first exemplary embodiment in that:
obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the normal pedestrian model, wherein the track similarity comprises the following steps:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a normal pedestrian model to the prediction positions D1, D2, …, Di, … and Dk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Si ^ e (-a × Di), wherein Di is the distance difference between the ith adoption time point Ti and the predicted position Di of the current time T0 and the current position G0 of the current time T0, and a is a prediction correction parameter;
according to the SUM { calcsore (Di, G0) × e (-Ti) | i ═ 1,2, …, k }, the trajectory similarity S based on the normal pedestrian model is obtained.
By implementing the second exemplary embodiment of the application, the trajectory similarity S based on the normal pedestrian model can be simply and quickly obtained, and the intelligence of the pedestrian trajectory prediction method is improved.
It should be noted that the value of the prediction correction parameter a is related to the long-term trajectory prediction result of the normal pedestrian model, and specifically includes: if the difference of the long-term track prediction results of the normal pedestrian model is large and the model prediction error is large, the value of a is smaller, otherwise, the value of a can be larger.
A third exemplary embodiment of the present application provides a pedestrian trajectory prediction method, which is a further improved solution based on the second exemplary embodiment of the present application, and the method is different from the second exemplary embodiment in that:
the obtaining of the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model includes:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a special pedestrian model to the prediction positions R1, R2, …, Ri, … and Rk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Ri ═ e (-b × Ri), wherein Ri is the distance difference between the ith adoption time point Ti and the predicted position Ri of the current time T0 and the current position G0 of the current time T0, and b is a predicted correction parameter;
according to the formula, the trajectory similarity R based on the special pedestrian model is obtained according to the formula of R { calcsore (Ri, G0) × e (-Ti) | i ═ 1,2, …, k }.
It should be noted that the value of the prediction correction parameter b is also related to the long-term trajectory prediction result of the special pedestrian model, specifically: if the difference of the long-term track prediction results of the special pedestrian models is large and the model prediction error is large, the value of b is smaller, otherwise, the value of b can be larger.
By implementing the third exemplary embodiment of the application, the trajectory similarity R based on the special pedestrian model can be simply and quickly obtained, and the intelligence of the pedestrian trajectory prediction method is further improved.
A fourth exemplary embodiment of the present application provides a pedestrian trajectory prediction method, which is a further improved solution based on the third exemplary embodiment of the present application, and the method is different from the third exemplary embodiment in that:
generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to a comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model, wherein the step of generating the target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model comprises the following steps:
acquiring the ratio of the track similarity S based on the normal pedestrian model to the track similarity R based on the special pedestrian model according to the track similarity S based on the normal pedestrian model and the track similarity R based on the special pedestrian model;
judging whether the ratio is smaller than a preset threshold value:
if yes, generating a target pedestrian predicted track based on the special pedestrian model;
and if not, generating a target pedestrian predicted track based on the normal pedestrian model.
By implementing the fourth exemplary embodiment of the present application, whether the ratio is smaller than a preset threshold is determined, and when the ratio is smaller than the preset threshold, the accuracy of the target pedestrian prediction trajectory based on the special pedestrian model to modify the automatic driving interaction strategy is generated, so that the probability of occurrence of a dangerous accident that a special pedestrian collides can be reduced to 5%.
It should be noted that the value range of the preset threshold is specifically 0.5 to 1.0.
Fifth to eighth exemplary embodiments of the present application provide a pedestrian trajectory prediction method, which is a further improved solution based on the first to fourth exemplary embodiments of the present application, respectively, and which differs from the first to fourth exemplary embodiments described above, respectively, in that:
before the actual track of the target pedestrian, the historical prediction track obtained based on the normal pedestrian model and the historical prediction track obtained based on the special pedestrian model are obtained, the method further comprises the following steps:
acquiring environmental information in a preset area;
extracting target pedestrian information in the environment information;
and acquiring the actual track of the target pedestrian according to the target pedestrian information.
By implementing the fifth to eighth exemplary embodiments of the present application, unnecessary noise information can be removed from the environment information to identify the actual trajectory of the target pedestrian, thereby further improving the accuracy of the pedestrian trajectory prediction method of the present application.
Fig. 2 shows a ninth exemplary embodiment of the present application, which provides a pedestrian trajectory prediction apparatus, including:
the track acquisition module is used for acquiring the actual track of the target pedestrian, the historical prediction track obtained based on the normal pedestrian model and the historical prediction track obtained based on the special pedestrian model;
the first processing module is used for obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the normal pedestrian model;
the second processing module is used for obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the special pedestrian model;
and the third processing module is used for generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
Further, the pedestrian trajectory prediction apparatus further includes:
the environment information acquisition module is used for acquiring environment information in a preset area;
the target pedestrian information extraction module is used for extracting target pedestrian information in the environment information;
and the actual track acquisition module of the target pedestrian is used for acquiring the actual track of the target pedestrian according to the information of the target pedestrian.
The modules of the pedestrian trajectory prediction device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of each functional unit or module is illustrated, and in practical applications, the above-mentioned function may be distributed as different functional units or modules as required, that is, the internal structure of the apparatus described in this application may be divided into different functional units or modules to implement all or part of the above-mentioned functions.
Fig. 3 illustrates a tenth exemplary embodiment of the present application providing an electronic device, which may be a server; the electronic device includes a processor, a memory, and a communication interface connected by a system bus. Wherein the processor of the device is configured to provide computing and control capabilities. The memory of the device may be implemented by any type or combination of volatile or non-volatile storage devices, including but not limited to: magnetic disk, optical disk, EEPROM, EPROM, SRAM, ROM, magnetic memory, flash memory, and PROM. The memory of the device provides an environment for the running of an operating system and computer programs stored within it. The communication interface of the device is a network interface for connecting and communicating with an external terminal through a network. The computer program, when executed by a processor, implements the steps of the pedestrian trajectory prediction method described in the above embodiments.
An eleventh exemplary embodiment of the present application provides a storage medium storing a computer program that, when executed by a processor, implements the pedestrian trajectory prediction method described in the above-described embodiment. Such storage media include, but are not limited to: ROM, RAM, CD-ROM, diskette, and floppy disk.
While the present embodiments have been described with reference to the accompanying drawings, it is to be understood that the invention is not limited to the precise embodiments described above, which are meant to be illustrative and not restrictive, and that various changes may be made therein by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (11)
1. A pedestrian trajectory prediction method is characterized by comprising the following steps:
acquiring an actual track of a target pedestrian, a historical predicted track obtained based on a normal pedestrian model and a historical predicted track obtained based on a special pedestrian model;
obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the normal pedestrian model;
obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track based on the special pedestrian model;
and generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
2. The method of claim 1, wherein the normal pedestrian model is a deep neural network model and the special pedestrian-based model is a regular logic model.
3. The method of claim 1, wherein the normal pedestrian model is a linear regression model and the special pedestrian-based model is a regular logic model.
4. The pedestrian trajectory prediction method according to claim 1, wherein the normal pedestrian model is an SVM model, and the special pedestrian-based model is a regular logic model.
5. The pedestrian trajectory prediction method according to any one of claims 2 to 4, wherein obtaining the similarity of the trajectories based on the normal pedestrian model according to the actual trajectory of the target pedestrian and the historical predicted trajectory based on the normal pedestrian model comprises:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a normal pedestrian model to the prediction positions D1, D2, …, Di, … and Dk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Si ^ e (-a × Di), wherein Di is the distance difference between the ith adoption time point Ti and the predicted position Di of the current time T0 and the current position G0 of the current time T0, and a is a prediction correction parameter;
according to the SUM { calcsore (Di, G0) × e (-Ti) | i ═ 1,2, …, k }, the trajectory similarity S based on the normal pedestrian model is obtained.
6. The pedestrian trajectory prediction method according to claim 5, wherein obtaining a similarity of the trajectory based on the special pedestrian model according to the actual trajectory of the target pedestrian and a historical predicted trajectory based on the special pedestrian model comprises:
acquiring a current position G0 of a current time T0 from an actual track of the target pedestrian;
respectively and correspondingly acquiring past k sampling time points T1, T2, …, Ti, … and Tk from a history prediction track of the target pedestrian based on a special pedestrian model to the prediction positions R1, R2, …, Ri, … and Rk of the current time T0, wherein i and k are natural numbers;
obtaining the position deviation of each predicted position and the current position according to Ri ═ e (-b × Ri), wherein Ri is the distance difference between the ith adoption time point Ti and the predicted position Ri of the current time T0 and the current position G0 of the current time T0, and b is a predicted correction parameter;
according to the formula, the trajectory similarity R based on the special pedestrian model is obtained according to the formula of R { calcsore (Ri, G0) × e (-Ti) | i ═ 1,2, …, k }.
7. The pedestrian trajectory prediction method according to claim 6, wherein generating the predicted trajectory of the target pedestrian based on the normal pedestrian model or the special pedestrian model according to a result of comparison of the similarity of the trajectory based on the normal pedestrian model and the similarity of the trajectory based on the special pedestrian model includes:
acquiring the ratio of the track similarity S based on the normal pedestrian model to the track similarity R based on the special pedestrian model according to the track similarity S based on the normal pedestrian model and the track similarity R based on the special pedestrian model;
judging whether the ratio is smaller than a preset threshold value:
if yes, generating a target pedestrian predicted track based on the special pedestrian model;
and if not, generating a target pedestrian predicted track based on the normal pedestrian model.
8. The pedestrian trajectory prediction method according to any one of claims 1 to 4, further comprising, before acquiring the actual trajectory of the target pedestrian, the historical predicted trajectory obtained based on the normal pedestrian model, and the historical predicted trajectory obtained based on the special pedestrian model:
acquiring environmental information in a preset area;
extracting target pedestrian information in the environment information;
and acquiring the actual track of the target pedestrian according to the target pedestrian information.
9. A pedestrian trajectory prediction device characterized by comprising:
the track acquisition module is used for acquiring the actual track of the target pedestrian, the historical prediction track obtained based on the normal pedestrian model and the historical prediction track obtained based on the special pedestrian model;
the first processing module is used for obtaining the track similarity based on the normal pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the normal pedestrian model;
the second processing module is used for obtaining the track similarity based on the special pedestrian model according to the actual track of the target pedestrian and the historical predicted track obtained based on the special pedestrian model;
and the third processing module is used for generating a target pedestrian predicted track based on the normal pedestrian model or the special pedestrian model according to the comparison result of the track similarity based on the normal pedestrian model and the track similarity based on the special pedestrian model.
10. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the pedestrian trajectory prediction method according to any one of claims 1 to 8 when executing the computer program.
11. A storage medium, characterized in that a computer program is stored in the storage medium, which computer program, when being executed by a processor, carries out the steps of the pedestrian trajectory prediction method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111475997.2A CN114312829B (en) | 2021-12-06 | 2021-12-06 | Pedestrian track prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111475997.2A CN114312829B (en) | 2021-12-06 | 2021-12-06 | Pedestrian track prediction method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114312829A true CN114312829A (en) | 2022-04-12 |
| CN114312829B CN114312829B (en) | 2024-04-23 |
Family
ID=81048151
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111475997.2A Active CN114312829B (en) | 2021-12-06 | 2021-12-06 | Pedestrian track prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114312829B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109029446A (en) * | 2018-06-22 | 2018-12-18 | 北京邮电大学 | A kind of pedestrian position prediction technique, device and equipment |
| CN110245609A (en) * | 2019-06-13 | 2019-09-17 | 深圳力维智联技术有限公司 | Pedestrian track generation method, device and readable storage medium storing program for executing |
| CN111797751A (en) * | 2020-06-29 | 2020-10-20 | 中国第一汽车股份有限公司 | Pedestrian trajectory prediction method, device, equipment and medium |
| WO2021062595A1 (en) * | 2019-09-30 | 2021-04-08 | Beijing Voyager Technology Co., Ltd. | Systems and methods for predicting a pedestrian movement trajectory |
| CN112805730A (en) * | 2019-12-30 | 2021-05-14 | 华为技术有限公司 | Trajectory prediction method and related equipment |
| US20210155266A1 (en) * | 2019-11-22 | 2021-05-27 | Samsung Electronics Co., Ltd. | System and method for object trajectory prediction in an autonomous scenario |
| CN113569980A (en) * | 2021-08-12 | 2021-10-29 | 中山大学 | Pedestrian movement track online prediction method and system in complex environment |
-
2021
- 2021-12-06 CN CN202111475997.2A patent/CN114312829B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109029446A (en) * | 2018-06-22 | 2018-12-18 | 北京邮电大学 | A kind of pedestrian position prediction technique, device and equipment |
| CN110245609A (en) * | 2019-06-13 | 2019-09-17 | 深圳力维智联技术有限公司 | Pedestrian track generation method, device and readable storage medium storing program for executing |
| WO2021062595A1 (en) * | 2019-09-30 | 2021-04-08 | Beijing Voyager Technology Co., Ltd. | Systems and methods for predicting a pedestrian movement trajectory |
| US20210155266A1 (en) * | 2019-11-22 | 2021-05-27 | Samsung Electronics Co., Ltd. | System and method for object trajectory prediction in an autonomous scenario |
| CN112805730A (en) * | 2019-12-30 | 2021-05-14 | 华为技术有限公司 | Trajectory prediction method and related equipment |
| CN111797751A (en) * | 2020-06-29 | 2020-10-20 | 中国第一汽车股份有限公司 | Pedestrian trajectory prediction method, device, equipment and medium |
| CN113569980A (en) * | 2021-08-12 | 2021-10-29 | 中山大学 | Pedestrian movement track online prediction method and system in complex environment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114312829B (en) | 2024-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111968375B (en) | Traffic flow prediction method and device, readable storage medium and electronic equipment | |
| CN112015843A (en) | Driving risk situation assessment method and system based on multi-vehicle intention interaction result | |
| CN113076599A (en) | Multimode vehicle trajectory prediction method based on long-time and short-time memory network | |
| JP2862153B2 (en) | Rule-based computer system, method of calling rules executed by the rule-based computer system, and incremental function learning method executed by the rule-based computer system | |
| CN110444011B (en) | Traffic flow peak identification method and device, electronic equipment and storage medium | |
| CN111007858A (en) | Training method of vehicle driving decision model, and driving decision determining method and device | |
| CN112733047B (en) | Vehicle foothold generation method, device, equipment and computer storage medium | |
| CN111238523A (en) | Method and device for predicting motion trail | |
| CN114881339A (en) | Vehicle trajectory prediction method, system, computer device, and storage medium | |
| CN111091215B (en) | Vehicle identification method, device, computer equipment and storage medium | |
| CN114021840A (en) | Channel switching strategy generation method and device, computer storage medium and electronic equipment | |
| Zhou et al. | Vision‐based approach for predicting the probability of vehicle–pedestrian collisions at intersections | |
| CN113674317B (en) | Vehicle tracking method and device for high-level video | |
| CN114312829A (en) | Pedestrian trajectory prediction method and device, electronic equipment and storage medium | |
| CN114118276A (en) | Network training method, control method and device | |
| Liu et al. | On‐road multi‐vehicle tracking algorithm based on an improved particle filter | |
| CN111126311B (en) | Urban road potential dangerous area identification method and device and electronic equipment | |
| CN117010272A (en) | Decision generation method, device, computer equipment and medium based on reinforcement learning | |
| CN111582589A (en) | Car rental insurance prediction method, device, equipment and storage medium | |
| CN114394099B (en) | Method and device for identifying abnormal running of vehicle, computer equipment and storage medium | |
| CN114596702B (en) | Traffic state prediction model construction method and traffic state prediction method | |
| CN113447037B (en) | Stroke yaw detection method and device | |
| Lee et al. | Sensor fusion for vehicle tracking based on the estimated probability | |
| CN117396389A (en) | Automatic driving instruction generation model optimization method, device, equipment and storage medium | |
| CN113177852A (en) | Intelligent policy pushing method, device, equipment and storage medium |
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 |