CN113096165B - Target object positioning method and device - Google Patents
Target object positioning method and device Download PDFInfo
- Publication number
- CN113096165B CN113096165B CN202110414088.1A CN202110414088A CN113096165B CN 113096165 B CN113096165 B CN 113096165B CN 202110414088 A CN202110414088 A CN 202110414088A CN 113096165 B CN113096165 B CN 113096165B
- Authority
- CN
- China
- Prior art keywords
- image
- target object
- acquiring
- image data
- coordinate
- 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
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/32—Determination of transform parameters for the alignment of images, i.e. image registration using correlation-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30221—Sports video; Sports image
- G06T2207/30228—Playing field
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30236—Traffic on road, railway or crossing
Abstract
The invention provides a target object positioning method and a target object positioning device, wherein the target object positioning method comprises the following steps: acquiring original road image data, wherein the original road image data comprises a target object; partitioning original road image data, and inputting the partitioned image into an image registration model; acquiring a deformation vector of each partitioned image according to the image registration model; acquiring a deformation field according to the deformation vector, and acquiring a registered image according to the deformation field and the blocked image; and positioning the target object according to the registered image. By implementing the method and the device, each camera does not need to be calibrated, and even if the original road image data is acquired through the uncalibrated camera, the registered image can be acquired through the image registration model, so that the target object can be accurately positioned.
Description
Technical Field
The invention relates to the field of traffic data processing, in particular to a target object positioning method and device.
Background
Tracking, trajectory prediction, and the like of a target pedestrian or a target vehicle require positioning of the target pedestrian or the target object first, and positioning of the target pedestrian or the target object is realized based on a picture taken by a camera mounted on a road. However, the photos taken by the camera have certain deformation, so that the camera can be calibrated before the camera is put into use in order to accurately position the target pedestrians and target objects, and the pedestrians or vehicles in the photos can be accurately positioned through the calibrated photos taken by the camera. However, with the continuous development of cities, more and more cameras are needed in roads, and the cameras need to be calibrated before each camera is installed, so that the workload is large.
Disclosure of Invention
Therefore, the technical problem to be solved by the present invention is to overcome the defect of the prior art that the workload for calibrating each camera to be used is large, and to provide a method and an apparatus for positioning a target object.
The invention provides a target object positioning method in a first aspect, which comprises the following steps: acquiring original road image data, wherein the original road image data comprises a target object; partitioning original road image data, and inputting the partitioned image into an image registration model; acquiring a deformation vector of each partitioned image according to the image registration model; acquiring a deformation field according to the deformation vector, and acquiring a registered image according to the deformation field and the blocked image; and positioning the target object according to the registered image.
Optionally, the target object positioning method provided by the present invention constructs an image registration model by the following steps: acquiring training image data, wherein the training image data comprises an input image and a real deformation vector corresponding to the input image; and inputting the input image and the real deformation vector corresponding to the input image into the neural network model for training to obtain an image registration model.
Optionally, the step of acquiring training image data includes: acquiring a first image through an uncalibrated camera; dicing the first image, and taking the diced image data as an input image; acquiring a standard image corresponding to the first image; and cutting the standard image into blocks, and taking image data corresponding to the input image in the cut image data as a real deformation vector of the input image.
Optionally, the step of acquiring training image data includes: acquiring a registered standard image; performing simulated deformation on the standard image according to the simulated deformation field to obtain a third image; cutting the third image into blocks, and taking the image data after cutting into blocks as an input image; and cutting the standard image into blocks, and taking an image corresponding to the input image in the image data after cutting as a real deformation vector of the input image.
Optionally, the step of locating the target object according to the registered image includes: acquiring image coordinates of the target object in the registered image; acquiring a first image coordinate of at least one characteristic point in the registered image, wherein the characteristic point is a static object serving as a reference standard in the registered image; acquiring a first map coordinate of the feature point in a map; acquiring a first transition matrix according to the first image coordinate and the first map coordinate; acquiring a first earth coordinate of the feature point according to the first map coordinate; acquiring a second transition matrix according to the first map coordinate and the first earth coordinate; and acquiring the earth coordinate of the target object according to the image coordinate, the first transition matrix and the second transition matrix.
Optionally, the target object positioning method provided by the present invention further includes: acquiring historical positioning information of a target object; acquiring a historical motion track of the target object according to the historical positioning information of the target object; and predicting the motion track of the target object at the next moment according to the positioning information of the target object and the historical motion track.
Optionally, the target object positioning method provided by the present invention further includes: acquiring the earth coordinates of a target object at different moments in a preset time period; and acquiring the moving speed of the target object in a preset time period according to the earth coordinates of the target object at different moments.
A second aspect of the present invention provides a target object positioning apparatus, including: the original road image data acquisition module is used for acquiring original road image data, and the original road image data comprises a target object; the system comprises an original road image data preprocessing module, an image registration module and a data processing module, wherein the original road image data preprocessing module is used for partitioning original road image data and inputting a partitioned image into an image registration model; the deformation vector acquisition module is used for acquiring the deformation vector of each blocked image according to the image registration model; the registered image acquisition module is used for acquiring a deformation field according to the deformation vector and acquiring a registered image according to the deformation field and the blocked image; and the target object positioning module is used for positioning the target object according to the registered image.
A third aspect of the present invention provides a computer apparatus comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executable by the at least one processor to perform the method of target object localization as provided by the first aspect of the present invention.
A fourth aspect of the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method for locating a target object as provided in the first aspect of the present invention.
The technical scheme of the invention has the following advantages:
1. the target object positioning method provided by the invention comprises the steps of partitioning original road image data after the original road image data are obtained, inputting the partitioned images into an image registration model, obtaining deformation vectors of each partitioned image according to the image registration model, obtaining the registered images according to the partitioned images and the respective deformation vectors thereof, and positioning the target object according to the registered images. According to the target object positioning method provided by the invention, each camera does not need to be calibrated, and even if the original road image data is acquired through the uncalibrated camera, the registered image can be acquired through the image registration model, so that the target object is accurately positioned.
2. The target object positioning method obtains the historical motion track of the target object according to the historical positioning information, and then predicts the motion track of the target object at the next moment according to the positioning information and the historical motion track of the target object. According to the target positioning method provided by the invention, the positioning information of the target object at different moments is obtained according to the image information obtained by different cameras, so that a historical motion track is formed, the track of the target object is predicted according to the historical motion track and the positioning information, the target object can be better tracked, and the road safety is guaranteed.
3. The target object positioning device provided by the invention is characterized in that after original road image data are obtained, the original road image data are firstly partitioned, then the partitioned images are input into the image registration model, deformation vectors of each partitioned image are obtained according to the image registration model, then the registered images are obtained according to the partitioned images and the respective deformation vectors thereof, and the target object is positioned according to the registered images. The target object positioning device provided by the invention does not need to calibrate each camera, and can obtain the registered image through the image registration model even if the original road image data is acquired through the uncalibrated camera, so that the target object is accurately positioned.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1-5 are flowcharts of specific examples of target location methods provided in embodiments of the present invention;
FIG. 6 is a functional block diagram of one particular example of a target locating device provided in an embodiment of the present invention;
fig. 7 is a functional block diagram of a computer device provided in an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. 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 invention.
In the description of the present invention, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Example 1
An embodiment of the present invention provides a target object positioning method, as shown in fig. 1, including:
step S110: and acquiring original road image data, wherein the original road image data comprises a target object. In a specific embodiment, the original road image data is acquired by cameras installed on two sides of a road, the cameras acquiring the original road image data can be unregistered cameras, if the acquired data is video data, image noise needs to be filtered by means of kalman filtering, particle filtering and the like, and then abnormal points of the image are eliminated by means of smooth filtering and the like. The target object may be a pedestrian, a vehicle, or the like.
Step S120: and partitioning the original road image data, and inputting the partitioned image into an image registration model. In one embodiment, the original road image data is first cut into blocks and then input into the image registration model because the image registration model has limited computational power.
Step S130: and acquiring the deformation vector of each partitioned image according to the image registration model.
Step S140: and acquiring a deformation field according to the deformation vector, and acquiring a registered image according to the deformation field and the blocked image. In a specific embodiment, the registered image can be obtained by interpolating the blocked image by using the deformation field.
Step S150: and positioning the target object according to the registered image.
According to the target object positioning method provided by the embodiment of the invention, after the original road image data is obtained, the original road image data is firstly partitioned, then the partitioned images are input into the image registration model, the deformation vector of each partitioned image is obtained according to the image registration model, then the registered images are obtained according to the partitioned images and the respective deformation vectors thereof, and the target object is positioned according to the registered images. According to the target object positioning method provided by the embodiment of the invention, each camera is not required to be calibrated, and even if the original road image data is acquired through the uncalibrated camera, the registered image can be acquired through the image registration model, so that the target object is accurately positioned.
In an alternative embodiment, as shown in fig. 2, the target object localization method provided by the embodiment of the present invention constructs an image registration model by the following steps:
step S210: training image data is obtained, wherein the training image data comprises an input image and a real deformation vector corresponding to the input image, the input image is image data with deformation, and the real deformation vector corresponding to the input image is image data without deformation corresponding to the input image.
Step S220: and inputting the input image and the real deformation vector corresponding to the input image into the neural network model for training to obtain an image registration model.
In an optional embodiment, in the step S210, the step of acquiring training image data specifically includes:
step S211: and acquiring a first image through the uncalibrated camera.
Step S212: in a specific embodiment, the image slicing refers to grid slicing of the image data, and the grid slicing of the first image is performed to obtain a plurality of pieces of image data, but when the neural network model is trained, only one piece of image data is used as input data for each training.
Step S213: in a specific embodiment, the standard image corresponding to the first image may be acquired by a calibrated camera, and parameters such as a position and an angle of the calibrated camera are adjusted, so that shooting parameters of the calibrated camera are consistent with parameters of the uncalibrated camera when acquiring the first image, and thus the standard image corresponding to the first image may be acquired by the calibrated camera.
Step S214: in a specific embodiment, when the standard image is diced, the size of the adopted grid is consistent with the size of the grid used when the first image is diced, and a deformation vector corresponding to the input image is obtained, namely, image data at the same position as the input image in the standard image is obtained.
In an alternative embodiment, unlike the steps S211 to S214, the step S210 includes:
step S215: acquiring a registered standard image, wherein in a specific embodiment, the registered image may be a standard image acquired by a calibrated camera, or may be an image data acquired by an uncalibrated camera, and then the image data is registered by a conventional classical registration method, and the registration is performed by a conventional classical registration method, although the registration speed is slow, the registration accuracy is high, and the conventional registration method is firstly registered by a method of calculating image gray level correlation, and then is matched by a spatial two-dimensional sliding template.
Step S216: and performing analog deformation on the standard image according to the analog deformation field to obtain a third image.
Step S217: the third image is diced, and the diced image data is used as an input image, which is described in detail in the above step S212.
Step S218: the standard image is cut into blocks, and the image corresponding to the input image in the cut image data is used as the true deformation vector of the input image, which is described in detail in the above step S214.
In an alternative embodiment, as shown in fig. 3, the step S150 specifically includes:
step S151: and acquiring the image coordinates of the target object in the registered image. In a specific embodiment, a target object is first selected through a target frame, and an image coordinate of the target object is a coordinate of a central point of the target frame in an image.
Step S152: and acquiring first image coordinates of at least one characteristic point in the registered image, wherein the characteristic point is a static object serving as a reference standard in the registered image, and can be a telegraph pole, a street lamp and the like.
Step S153: and acquiring first map coordinates of the feature points in the map.
Step S154: acquiring a first transition matrix according to the linear transformation of the first image coordinate and the first map coordinate;
step S155: acquiring a first earth coordinate of the feature point according to the first map coordinate;
step S156: acquiring a second transition matrix according to the first map coordinate and the first earth coordinate;
step S157: and acquiring the earth coordinate of the target object according to the image coordinate, the first transition matrix and the second transition matrix. After a first transition matrix between the image coordinate and the map coordinate and a second transition matrix between the map coordinate and the earth are obtained, the image coordinate of the target object can be converted into the earth coordinate by utilizing linear transformation, and the target object can be positioned.
In a specific embodiment, as shown in fig. 4, the target object positioning method provided in the embodiment of the present invention further includes:
step S160: and acquiring historical positioning information of the target object. The historical positioning information of the target object is obtained through pictures which are taken by the camera devices at different positions at different moments.
Step S170: and acquiring the historical motion trail of the target object according to the historical positioning information of the target object. For example, the historical motion trajectory may be a motion trajectory of the target object over a certain period of time within a month.
Step S180: and predicting the motion track of the target object at the next moment according to the positioning information of the target object and the historical motion track. In a specific embodiment, when a target object is found in a road, the travel trajectory of the target object at this time can be predicted according to the historical trajectory of the target object in the time period, and then the motion trajectory of the target object at the next time is determined according to the positioning of the target object and the predicted trajectory.
According to the target object positioning method provided by the embodiment of the invention, the historical motion track of the target object is obtained according to the historical positioning information, and then the motion track of the target object at the next moment is predicted according to the positioning information and the historical motion track of the target object. According to the target positioning method provided by the embodiment of the invention, the positioning information of the target object at different moments is obtained according to the image information obtained by different cameras, so that a historical motion track is formed, the track of the target object is predicted according to the historical motion track and the positioning information, the target object can be better tracked, and the road safety can be further ensured by predicting the track of each object in the road.
In a specific embodiment, the method for predicting the motion trajectory of the target object in steps S160-S180 is applicable to both vehicles and pedestrians, and for the trajectory prediction of the vehicle, the lane where the target vehicle is located may be determined first, and if the target vehicle is in a right-turn lane, the target vehicle is considered to turn right at the next intersection. According to the method, the motion track of the target object at the next moment is predicted only through one picture without the need of based on the historical motion track of the target vehicle, but the prediction accuracy is possibly low.
In a specific embodiment, as shown in fig. 5, the target object positioning method provided in the embodiment of the present invention further includes:
step S190: acquiring the earth coordinates of a target object at different moments in a preset time period;
step S111: and acquiring the moving speed of the target object in a preset time period according to the earth coordinates of the target object at different moments.
In a specific embodiment, if the preset time period is shorter, the two images continuously acquired by the same camera both include the target object, and the moving speed of the target object can be directly acquired through image coordinates without being converted into earth coordinates.
Example 2
An embodiment of the present invention provides a target object positioning apparatus, as shown in fig. 6, including:
the original road image data obtaining module 110 is configured to obtain original road image data, where the original road image data includes a target object, and the detailed description is described in the foregoing embodiment 1 for step S110.
The original road image data preprocessing module 120 is configured to block original road image data, and input the blocked image to the image registration model, which is described in detail in embodiment 1 above for step S120.
A deformation vector obtaining module 130, configured to obtain a deformation vector of each segmented image according to the image registration model, which is described in detail in the above embodiment 1 for the description of step S130.
The registered image obtaining module 140 is configured to obtain a deformation field according to the deformation vector, and obtain a registered image according to the deformation field and the blocked image, which is described in detail in the above embodiment 1 for the step S140.
The target object positioning module 150 is configured to position the target object according to the registered image, and the detailed description is described in the above embodiment 1 for the step S150.
According to the target object positioning device provided by the embodiment of the invention, after the original road image data is obtained, the original road image data is firstly partitioned, then the partitioned images are input into the image registration model, the deformation vector of each partitioned image is obtained according to the image registration model, then the registered images are obtained according to the partitioned images and the respective deformation vectors thereof, and the target object is positioned according to the registered images. The target object positioning device provided by the embodiment of the invention does not need to calibrate each camera, and can obtain the registered image through the image registration model even if the original road image data is obtained through the uncalibrated camera, so that the target object is accurately positioned.
Example 3
An embodiment of the present invention provides a mobile terminal device, as shown in fig. 7, the computer device may include a processor 71 and a memory 72, where the processor 71 and the memory 72 may be connected by a bus or in another manner, and fig. 7 takes the connection by the bus as an example.
The processor 71 may be a Central Processing Unit (CPU). The Processor 71 may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, or combinations thereof.
The memory 72, which is a non-transitory computer readable storage medium, may be used for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the target object locating method in the embodiment of the present invention. The processor 71 executes various functional applications and data processing of the processor by executing non-transitory software programs, instructions and modules stored in the memory 72, namely, implements the target object locating method in the above method embodiment.
The memory 72 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created by the processor 71, and the like. Further, the memory 72 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 72 may optionally include memory located remotely from the processor 71, and such remote memory may be connected to the processor 71 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
One or more modules are stored in the memory 72 and, when executed by the processor 71, perform the target object localization method in the embodiments shown in fig. 1-5.
Example 4
The present invention provides a computer readable storage medium, which stores computer instructions, and it will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program to instruct related hardware, and the program can be stored in a computer readable storage medium, and when executed, the program can include the processes of the embodiments of the methods. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk (Hard Disk Drive, abbreviated as HDD) or a Solid State Drive (SSD), etc.; the storage medium may also comprise a combination of memories of the kind described above.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (9)
1. A method for locating a target object, comprising:
acquiring original road image data, wherein the original road image data comprises a target object;
partitioning the original road image data, and inputting the partitioned image into an image registration model;
acquiring a deformation vector of each partitioned image according to the image registration model;
acquiring a deformation field according to the deformation vector, and acquiring a registered image according to the deformation field and the blocked image;
positioning the target object according to the registered image;
the step of locating the target object according to the registered image comprises:
acquiring image coordinates of the target object in the registered image;
acquiring a first image coordinate of at least one characteristic point in the registered image, wherein the characteristic point is a static object serving as a reference standard in the registered image;
acquiring a first map coordinate of the feature point in a map;
acquiring a first transition matrix according to the first image coordinate and the first map coordinate;
acquiring a first earth coordinate of the feature point according to the first map coordinate;
acquiring a second transition matrix according to the first map coordinate and the first earth coordinate;
and acquiring the earth coordinate of the target object according to the image coordinate, the first transition matrix and the second transition matrix.
2. The target object localization method according to claim 1, wherein the image registration model is constructed by:
acquiring training image data, wherein the training image data comprises an input image and a real deformation vector corresponding to the input image;
and inputting the input image and the real deformation vector corresponding to the input image into a neural network model for training to obtain the image registration model.
3. The method of claim 2, wherein the step of acquiring training image data comprises:
acquiring a first image through an uncalibrated camera;
dicing the first image, and taking the diced image data as the input image;
acquiring a standard image corresponding to the first image;
and cutting the standard image into blocks, and taking image data corresponding to the input image in the cut image data as a real deformation vector of the input image.
4. The method of claim 2, wherein the step of acquiring training image data comprises:
acquiring a registered standard image;
performing simulated deformation on the standard image according to the simulated deformation field to obtain a third image;
dicing the third image, and taking the diced image data as the input image;
and cutting the standard image into blocks, and taking an image corresponding to the input image in the image data after cutting as a real deformation vector of the input image.
5. The target object positioning method according to claim 1, further comprising:
acquiring historical positioning information of the target object;
acquiring a historical motion track of the target object according to the historical positioning information of the target object;
and predicting the motion trail of the target object at the next moment according to the positioning information of the target object and the historical motion trail.
6. The target object positioning method according to claim 1, further comprising:
acquiring the earth coordinates of the target object at different moments in a preset time period;
and acquiring the moving speed of the target object in the preset time period according to the earth coordinates of the target object at different moments.
7. A target object positioning apparatus, comprising:
the system comprises an original road image data acquisition module, a target object acquisition module and a target object acquisition module, wherein the original road image data acquisition module is used for acquiring original road image data which comprises the target object;
the system comprises an original road image data preprocessing module, an image registration module and a data processing module, wherein the original road image data preprocessing module is used for partitioning the original road image data and inputting the partitioned image into an image registration model;
a deformation vector obtaining module, configured to obtain a deformation vector of each segmented image according to the image registration model;
a registered image obtaining module, configured to obtain a deformation field according to the deformation vector, and obtain a registered image according to the deformation field and the blocked image;
the target object positioning module is used for positioning the target object according to the registered image;
the locating the target object according to the registered image comprises:
acquiring image coordinates of the target object in the registered image;
acquiring a first image coordinate of at least one characteristic point in the registered image, wherein the characteristic point is a static object serving as a reference standard in the registered image;
acquiring a first map coordinate of the feature point in a map;
acquiring a first transition matrix according to the first image coordinate and the first map coordinate;
acquiring a first earth coordinate of the feature point according to the first map coordinate;
acquiring a second transition matrix according to the first map coordinate and the first earth coordinate;
and acquiring the earth coordinate of the target object according to the image coordinate, the first transition matrix and the second transition matrix.
8. A computer device, comprising:
at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to perform the target object localization method of any of claims 1-6.
9. A computer-readable storage medium storing computer instructions for causing a computer to perform the target object localization method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110414088.1A CN113096165B (en) | 2021-04-16 | 2021-04-16 | Target object positioning method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110414088.1A CN113096165B (en) | 2021-04-16 | 2021-04-16 | Target object positioning method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113096165A CN113096165A (en) | 2021-07-09 |
| CN113096165B true CN113096165B (en) | 2022-02-18 |
Family
ID=76678533
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110414088.1A Active CN113096165B (en) | 2021-04-16 | 2021-04-16 | Target object positioning method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113096165B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105913459A (en) * | 2016-05-10 | 2016-08-31 | 中国科学院自动化研究所 | Moving object detection method based on high resolution continuous shooting images |
| CN107590832A (en) * | 2017-09-29 | 2018-01-16 | 西北工业大学 | Physical object tracking positioning method based on physical feature |
| CN108108697A (en) * | 2017-12-25 | 2018-06-01 | 中国电子科技集团公司第五十四研究所 | A kind of real-time UAV Video object detecting and tracking method |
| CN111091111A (en) * | 2019-12-25 | 2020-05-01 | 河北工业大学 | Vehicle bottom dangerous target identification method |
| CN111524169A (en) * | 2019-01-25 | 2020-08-11 | 罗伯特·博世有限公司 | Localization based on image registration of sensor data and map data with neural networks |
| CN111612841A (en) * | 2020-06-22 | 2020-09-01 | 上海木木聚枞机器人科技有限公司 | Target positioning method and device, mobile robot and readable storage medium |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7856120B2 (en) * | 2007-03-30 | 2010-12-21 | Mitsubishi Electric Research Laboratories, Inc. | Jointly registering images while tracking moving objects with moving cameras |
| CN104111456B (en) * | 2014-07-23 | 2017-06-06 | 中国国土资源航空物探遥感中心 | A kind of line of high-speed railway Ground Deformation high-resolution InSAR monitoring methods |
| US10102640B2 (en) * | 2016-11-29 | 2018-10-16 | Optinav Sp. Z O.O. | Registering three-dimensional image data of an imaged object with a set of two-dimensional projection images of the object |
| CN110491156A (en) * | 2019-08-27 | 2019-11-22 | 无锡物联网创新中心有限公司 | A kind of cognitive method, apparatus and system |
| CN111047629B (en) * | 2019-11-04 | 2022-04-26 | 中国科学院深圳先进技术研究院 | Multi-modal image registration method and device, electronic equipment and storage medium |
| CN112365404B (en) * | 2020-11-23 | 2023-03-17 | 成都唐源电气股份有限公司 | Contact net panoramic image splicing method, system and equipment based on multiple cameras |
-
2021
- 2021-04-16 CN CN202110414088.1A patent/CN113096165B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105913459A (en) * | 2016-05-10 | 2016-08-31 | 中国科学院自动化研究所 | Moving object detection method based on high resolution continuous shooting images |
| CN107590832A (en) * | 2017-09-29 | 2018-01-16 | 西北工业大学 | Physical object tracking positioning method based on physical feature |
| CN108108697A (en) * | 2017-12-25 | 2018-06-01 | 中国电子科技集团公司第五十四研究所 | A kind of real-time UAV Video object detecting and tracking method |
| CN111524169A (en) * | 2019-01-25 | 2020-08-11 | 罗伯特·博世有限公司 | Localization based on image registration of sensor data and map data with neural networks |
| CN111091111A (en) * | 2019-12-25 | 2020-05-01 | 河北工业大学 | Vehicle bottom dangerous target identification method |
| CN111612841A (en) * | 2020-06-22 | 2020-09-01 | 上海木木聚枞机器人科技有限公司 | Target positioning method and device, mobile robot and readable storage medium |
Non-Patent Citations (4)
| Title |
|---|
| Modeling lung deformation: A combined deformable image registration method with spatially varying Young"s modulus estimates;Min Li等;《Medical physics》;20131231;第40卷(第8期);第1-11页 * |
| 全局图像配准的目标快速定位方法;王志强等;《红外与激光工程》;20151231;第44卷(第S期);第2节 * |
| 基于深度学习的医学图像配准综述;zlinzju;《https://blog.csdn.net/weixin_43026262/article/details/102564308》;20191015;正文第2页 * |
| 视频图像中运动目标检测与跟踪方法研究;天鹏辉;《中国博士学位论文全文数据库 (信息科技辑)》;20140615;第I138-35页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113096165A (en) | 2021-07-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109215433B (en) | Vision-based driving scenario generator for automated driving simulation | |
| JP6494719B2 (en) | Traffic signal map creation and detection | |
| US11487988B2 (en) | Augmenting real sensor recordings with simulated sensor data | |
| EP3836018B1 (en) | Method and apparatus for determining road information data and computer storage medium | |
| CN112417953B (en) | Road condition detection and map data updating method, device, system and equipment | |
| EP3211596A1 (en) | Generating a virtual world to assess real-world video analysis performance | |
| CN111771206A (en) | Map zoning system for autonomous vehicles | |
| CN112105890A (en) | RGB point cloud based map generation system for autonomous vehicles | |
| CN111771229A (en) | Point cloud ghost effect detection system for automatic driving vehicle | |
| CN112055805A (en) | Point cloud registration system for autonomous vehicles | |
| KR20180046798A (en) | Method and apparatus for real time traffic information provision | |
| US20200143179A1 (en) | Infrastructure-free nlos obstacle detection for autonomous cars | |
| CN111220993A (en) | Target scene positioning method and device, computer equipment and storage medium | |
| CN111860227A (en) | Method, apparatus, and computer storage medium for training trajectory planning model | |
| US20200279395A1 (en) | Method and system for enhanced sensing capabilities for vehicles | |
| Espejel-García et al. | An alternative vehicle counting tool using the Kalman filter within MATLAB | |
| CN115019060A (en) | Target recognition method, and training method and device of target recognition model | |
| CN113012215A (en) | Method, system and equipment for space positioning | |
| CN114299230A (en) | Data generation method and device, electronic equipment and storage medium | |
| US20210323577A1 (en) | Methods and systems for managing an automated driving system of a vehicle | |
| CN113096165B (en) | Target object positioning method and device | |
| CN110781730A (en) | Intelligent driving sensing method and sensing device | |
| US11908095B2 (en) | 2-D image reconstruction in a 3-D simulation | |
| CN112433193B (en) | Multi-sensor-based mold position positioning method and system | |
| CN110634157B (en) | Method for determining transformation matrix |
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 |