WO2020172870A1 - 一种目标对象的移动轨迹确定方法和装置 - Google Patents

一种目标对象的移动轨迹确定方法和装置 Download PDF

Info

Publication number
WO2020172870A1
WO2020172870A1 PCT/CN2019/076557 CN2019076557W WO2020172870A1 WO 2020172870 A1 WO2020172870 A1 WO 2020172870A1 CN 2019076557 W CN2019076557 W CN 2019076557W WO 2020172870 A1 WO2020172870 A1 WO 2020172870A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
target object
target
determining
stored
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.)
Ceased
Application number
PCT/CN2019/076557
Other languages
English (en)
French (fr)
Inventor
熊风
庞磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
SZ DJI Technology Co Ltd
Original Assignee
SZ DJI Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by SZ DJI Technology Co Ltd filed Critical SZ DJI Technology Co Ltd
Priority to PCT/CN2019/076557 priority Critical patent/WO2020172870A1/zh
Priority to CN201980005227.6A priority patent/CN111263955A/zh
Publication of WO2020172870A1 publication Critical patent/WO2020172870A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/292—Multi-camera tracking
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/61—Control of cameras or camera modules based on recognised objects
    • H04N23/611—Control of cameras or camera modules based on recognised objects where the recognised objects include parts of the human body
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/695—Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects

Definitions

  • the present invention relates to the technical field of image processing, and in particular to a method and device for determining the movement track of a target object.
  • Cameras are widely used in the surveillance field.
  • the monitoring personnel monitor the target object by analyzing the images captured by the camera.
  • the surveillance personnel analyze the images output by the camera that captures the target object one by one to realize the analysis of the movement trajectory of the target object, which is low in efficiency, high in work intensity, and low in accuracy.
  • the embodiment of the present invention provides a method and device for determining the movement track of a target object, so as to determine the movement track of the target object through images taken by multiple photographing devices.
  • the first aspect of the embodiments of the present invention provides a method for determining the movement track of a target object, including:
  • the target image is determined from the images stored in the storage device, wherein the image stored in the storage device includes images of objects photographed by multiple photographing devices, and the multiple photographing devices include at least two photographing devices arranged at different positions.
  • the target image includes the image of the target object in the image of the object;
  • the movement track of the target object is determined according to the target image.
  • a device for determining a movement track of a target object including:
  • the memory is used to store program codes
  • the processor calls the program code, and when the program code is executed, is used to perform the following operations:
  • the target image is determined from the images stored in the storage device, wherein the image stored in the storage device includes images of objects photographed by multiple photographing devices, and the multiple photographing devices include at least two photographing devices arranged at different positions.
  • the target image includes the image of the target object in the image of the object;
  • the movement track of the target object is determined according to the target image.
  • the method and device for determining the movement trajectory of a target object provided by the embodiments of the present invention associate images of the target object captured by multiple shooting devices, and determine the movement trajectory of the target object through the images of the target object captured by the multiple shooting devices. Greatly reduce the workload of monitoring and improve accuracy.
  • FIG. 1 is a schematic flowchart of a method for determining a movement track of a target object according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram of a scene for determining the movement track of a target object provided by an embodiment of the present invention
  • FIG. 3 is a schematic diagram showing the movement track of a target object according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of determining the position of a target object provided by an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of determining the orientation of the target object relative to the shooting device according to an embodiment of the present invention.
  • Fig. 6 is a schematic diagram of deleting a background image in a bounding box provided by an embodiment of the present invention.
  • Fig. 7 is a structural diagram of a device for determining a movement track of a target object provided by an embodiment of the present invention.
  • FIG. 1 shows a schematic flowchart of a method for determining a movement track of a target object according to an embodiment of the present invention, where the method may be executed by an apparatus for determining a movement track of a target object.
  • an embodiment of the present invention provides a method for determining a movement track of a target object, and the method 100 includes:
  • the target image is determined from the images stored in the storage device, wherein the image stored in the storage device includes images of objects photographed by multiple photographing devices, and the multiple photographing devices include at least two images set in different positions.
  • a photographing device, the target image includes an image of the target object in the image of the object;
  • a monitoring area 201 (such as a shopping mall, an office building, etc.) may be provided with multiple photographing devices 201 (for example, including at least two photographing devices), wherein each photographing device may be set at a different position of the monitoring area 201, and each One photographing device 201 can photograph one photographing area (for example, a fan-shaped area as shown in the figure) to obtain an image.
  • the image may be stored in a storage device.
  • the server may obtain images captured by the multiple photographing devices 201 and store the images in a storage device.
  • the storage device may be a local storage device of the server, or may be a storage device communicatively connected with the server.
  • the object 203 When the object 203 moves in the monitoring area 201, the object 203 will be photographed by multiple photographing devices 201, that is, the photographing device 201 will output an image of the object, that is, the storage device 201 will store the image of the object in this way.
  • the object 203 may be any object that can move autonomously or move by receiving external forces.
  • the object may be a pedestrian, an animal, a robot, etc., and the object is a pedestrian for schematic illustration here.
  • the object includes the target object, and the target object is an interested or designated object among the objects.
  • the storage device stores the image of the target object, that is, the target image.
  • the device for determining the movement trajectory of the target object can determine the image of the target object from the images stored in the storage device, that is, the target image can be obtained by filtering the images stored in the storage device.
  • step S120 the movement track of the target object is determined according to the target image.
  • the target image is associated with the photographing device 201, and the photographing device 201 is set in a different position of the monitoring area 201, the target image can be used to indicate that the target object appears in the monitoring area 201 s position.
  • the device for determining the movement track of the target object can determine the movement track 204 of the target object in the monitoring area 201 according to the target image.
  • the device for determining the movement track of the target object may be a device that communicates with the storage device, such as a computer.
  • the device for determining the movement track of the target object may be the server.
  • the images of the target object captured by multiple photographing devices are associated, and the movement trajectory of the target object is determined through the images of the target object photographed by the multiple photographing devices. Reduce the workload of monitoring and improve accuracy.
  • the method further includes: displaying the movement track on an interactive interface.
  • the device for determining the movement track of the target object may display the movement track on the interactive interface.
  • the interactive interface may be a display screen.
  • the display of the movement track can be displayed in the form of a line drawing of the track, for example, by displaying the line drawing of the moving track 204 shown in FIG. 2; or in the form of text, such as
  • the display is shown in the manner shown in FIG. 3, for example, the movement track is displayed in the form of displaying the position of the target object according to time. It is understandable that those skilled in the art can also display the movement track in any other manner that can display the time and the corresponding position of the target object, which is not limited herein.
  • the determining the target image from the images stored in the storage device includes: matching the reference image of the target object with the image stored in the storage device to determine the target image from the images stored in the storage device.
  • the device for determining the movement trajectory of the target object may obtain the reference image of the target object, and then compare the reference image of the target object with the image stored in the storage device.
  • a matching operation is performed to determine a target image from the stored images, and further, a matching operation is performed on the reference image of the target object and the image of the object stored in the storage device to determine the target from the stored image of the object image.
  • the reference image of the target object may be stored in the local storage device of the device for determining the movement trajectory of the target object, and the reference image of the target object may also be stored in a storage device communicatively connected with the device for determining the movement trajectory of the target object, The device for determining the movement track of the target object may obtain the reference image of the target object from the storage device.
  • the matching the reference image of the target object with the image stored in the storage device to determine the target image from the images stored in the storage device includes: comparing the reference image of the target object with the image stored in the storage device Matching is performed to determine the matching coefficient between the reference image and the stored image; when the matching coefficient satisfies a preset condition, the image corresponding to the matching coefficient in the stored images is determined as the target image.
  • the device for determining the movement trajectory of the target object may match the reference image of the target object with the stored image, and the matching coefficient may be determined through the matching, wherein the matching coefficient may be a characteristic of the stored image. A parameter of the degree of similarity between the image and the reference image.
  • the matching coefficient meets a preset condition, it is determined that the stored image is the target image, and when the matching coefficient does not meet the preset condition, it is determined that the stored image is not the target image.
  • the matching coefficient includes the Euclidean distance between the stored image and the reference image.
  • the smaller the Euclidean distance the higher the similarity between the stored image and the reference image.
  • the Euclidean distance between the stored image and the reference image is less than the preset Euclidean geometric distance threshold, it means The stored image is the target image.
  • the Euclidean distance between the stored image and the reference image is greater than or equal to the preset Euclidean geometric distance threshold, it is determined that the stored image is not the target image.
  • the device for determining the movement trajectory of the target object can extract features from the stored image and the reference image through the trained deep learning model, where the feature is a feature vector, so that the feature X of the stored image can be obtained as (x 1 , X 2 ,...x n ) and the feature Y of the reference image is (y 1 , y 2 ,...y n ), then the Euclidean distance between the stored image and the reference image
  • the extracted features can use various appropriate feature extraction methods such as LBP (local binary mode), HoG (directional gradient histogram), PCA (principal component analysis) or neural network for feature extraction and The multiple feature vectors are generated.
  • LBP local binary mode
  • HoG directional gradient histogram
  • PCA principal component analysis
  • neural network for feature extraction and The multiple feature vectors are generated.
  • the same feature extraction method is used for the face of the object in each frame of the reference image of the target object and/or the image of the object stored in the storage device Generate feature vectors.
  • feature extraction can be performed on the face image area as a whole, or feature extraction can be performed separately in different sub-image areas of the face image area.
  • the feature vector of the reference image and the feature vector of the image stored in the storage device may be normalized vectors, that is, the vectors are normalized so that the modulus of the vector is 1, which is beneficial to improve the calculation efficiency. It is convenient to quickly detect the target image.
  • the method further includes: acquiring index information of the target object input by the user through an interactive interface, and determining the reference image of the target object according to the index information of the target object.
  • the user when the user views the movement track of a specific target object, the user can input index information of the target object on the interactive interface, where the index information can be any information that can index the target object.
  • the index information of the target object has a corresponding relationship with the reference image of the target object, and the device for determining the movement track of the target object may determine the reference image of the target object according to the index information of the target object.
  • the index information may be at least one of identity information, biometric information, and images of the target object.
  • the identity information includes at least one of employee ID, ID number, name, insurance account number, etc.
  • the biometric information may include at least one of fingerprint, voiceprint, iris, etc.
  • the action may include the reference image of the target object as described above.
  • the index information includes the identity information of the target object, the obtaining the index information of the target object input by the user through the interactive interface, and determining the reference image of the target object according to the index information of the target object includes: obtaining the user passed The identity information of the target object input by the interactive interface; the image corresponding to the identity information is obtained according to the identity information; the image corresponding to the identity information is determined as the reference image of the target object.
  • the index information includes an image
  • the acquiring index information of the target object input by the user through the interactive interface, and determining the reference image of the target object according to the index information of the target object includes: acquiring the user input through the interactive interface The image of the target object, and the input image is determined as the reference image of the target object.
  • the image may be an image output by at least one of the multiple photographing devices, or an image of a target object obtained by other means.
  • the device for determining the movement trajectory of the target object controls the interactive interface to display images captured and displayed by multiple shooting devices, and determines the user from the displayed image according to the selection operation input by the user received. For the selected image, the image selected by the user is determined as the reference image of the target object. In this way, the user can select and determine the image including the target object output by the photographing device as the reference image of the target object.
  • the determining the movement track of the target object according to the target image includes: determining a photographing device that takes the target image according to the target image; and determining the target object according to the installation position of the photographing device that takes the target image The location information of the target object; the movement track of the target object according to the location information of the target object.
  • the device for determining the movement trajectory of the target object may determine the shooting device that took the target image according to the target image and the corresponding relationship, and obtain the The installation position of the photographing device may further determine the position information of the target object according to the installation position of the photographing device. After the location information of the target object is determined, the movement track of the target object can be determined according to the location information of the target object. In some embodiments, when the accuracy requirement is not high, the installation position of the shooting device of the target image may be determined as the position of the target object.
  • the determining the position of the target object according to the installation position of the shooting device that shoots the target image includes: according to the installation position and shooting direction of the shooting device that shoots the target image, and when the shooting device is shooting The position of the target object in the shooting frame in the target image determines the position of the target object.
  • the photographing device 401 photographs and outputs a target image of the target object 402, and the distance d between the target object 402 and the photographing device can be determined according to the target image, for example, the image area corresponding to the target object in the target image ( The size of the target image may be used to determine the distance d.
  • the size of the target image may be used to determine the distance d.
  • the image area is large, it means that the distance d between the target object 402 and the photographing device is relatively close, and when the image area is small, it means that the distance d between the target object 402 and the photographing device is relatively far.
  • the device for determining the movement track of the target object may determine the orientation of the target object relative to the shooting device according to the shooting direction of the shooting device and the position of the target object in the shooting screen when the shooting device shoots the target image.
  • the position of the target object in the shooting frame 501 is o 1
  • the deviation between the position o 1 and the center o of the shooting frame 501 can be determined, and the deviation can include the deviation x in the horizontal direction and the vertical The deviation y in the straight direction.
  • the deviation may reflect the angle at which the target object deviates from the shooting direction of the shooting device, and the angle at which the target object deviates from the shooting direction of the shooting device can be determined according to the deviation, and then the target object is determined according to the shooting direction and the angle of the shooting device
  • the orientation relative to the camera may include at least a pitch azimuth and a yaw azimuth.
  • the device for determining the movement track of the target object may determine the position of the target object relative to the photographing device according to the distance d and the orientation, and further may determine the position of the target object relative to the photographing device according to the position of the target object relative to the photographing device and the The installation location of the camera determines the location of the target object.
  • the method further includes: acquiring images output by the plurality of photographing devices; determining an image of the object from the output images according to an object recognition algorithm; storing the image of the object in the storage device .
  • the device for determining the movement trajectory of the target object can acquire the image output by the photographing device.
  • the device for determining the movement trajectory of the target object may run an object recognition algorithm on the image for each acquired frame of image or periodically at a preset interval, and in this way determine the image of the object from the image.
  • the determining the image of the object from the output image according to the object recognition algorithm includes: determining the bounding box of the object in the output image according to the object recognition algorithm; and determining the image of the object according to the bounding box .
  • the bounding box 602 of the object in the image can be determined from the image through the recognition algorithm, for example, from the image through a neural network. Determine the bounding box of the object in the image. Further, the image of the object is determined according to the bounding box. In some embodiments, the image within the bounding box may be determined as the image of the object.
  • the determining the image of the object according to the bounding box includes: identifying the image of the background from the image in the bounding box, deleting the image of the background; and determining the remaining image in the bounding box as the image of the object image.
  • the device for determining the movement trajectory of the target object can identify the background image 603 in the bounding box 602, Delete the background image 603. In this way, after the deletion is completed, the remaining image 604 in the bounding box 602 can be determined as the target image.
  • the object is a pedestrian
  • the method further includes: determining an image including the front face of the target object from the target image, and displaying the image including the front face of the target object.
  • the device for determining the movement track of the target object determines an image including the front face of the target object from the target image, and controls the interactive interface to display the image of the front face of the target object. Further, the image area of the front face can be determined from the image of the front face of the target object, the image in the image area is enlarged, and then the interactive interface is controlled to display the enlarged image to identify the target more clearly Pedestrian's face.
  • an image that includes the front face of the target object may be determined from the target image by a face quality evaluation algorithm, where the face quality evaluation algorithm may obtain the profile of the face image in the pedestrian sequence based on a neural network model. The degree of blur and the angle of the face, and different weights are set to evaluate the quality of the face picture.
  • the degree of blur of the human face picture may be a neighborhood gradient change of the human face picture and the blurred image of the human face picture.
  • the texture of a clear image is sharp, and the neighborhood of the image changes significantly; the neighborhood of a blurred image changes little.
  • the face picture is passed through a preset low-pass filter to obtain a blurred image of the face picture, and the neighborhood gradient change of the blurred image is calculated; by comparing the selected face The neighborhood gradient of the picture and the blurred image is changed to obtain the degree of blur of the face picture.
  • the angle of the face can be calculated by deep learning based on the convolutional neural network structure to obtain the angle of the face, or the angle of the face can be regression estimated by the adaptive random forest algorithm.
  • the method further includes: performing recognition according to the image including the front face of the target object to determine whether the target object belongs to a preset organization; when it is determined that the target object belongs to the preset organization, displaying the A prompt message, wherein the first prompt information is used to indicate that the target object belongs to a preset organization; when it is determined that the target object does not belong to the preset organization, the second prompt information is displayed, wherein the second prompt The information is used to indicate that the target row does not belong to the preset organization.
  • the device for determining the movement track of the target object determines whether the target object belongs to a preset organization, for example, determines whether the target pedestrian belongs to a certain company, department, group, owner group, etc., when it is determined that the target object belongs to the preset organization
  • the first prompt information is displayed, and when it is determined that the target object does not belong to the preset organization, the second prompt information is displayed to realize the effect of an alarm.
  • the target object is a pedestrian.
  • Performing face recognition based on the frontal image to confirm the identity information of the target object further improves the accuracy of pedestrian recognition and improves the recognition efficiency.
  • FIG. 7 shows a schematic block diagram of an apparatus for determining a movement track of a target object according to an embodiment of the present invention.
  • an embodiment of the present invention provides a device for determining a movement track of a target object, and the device 700 includes:
  • Memory 701 and processor 702 are Memory 701 and processor 702;
  • the memory is used to store program codes
  • the processor 702 calls the program code, and when the program code is executed, is used to perform the following operations:
  • the target image is determined from the images stored in the storage device, wherein the image stored in the storage device includes images of objects photographed by multiple photographing devices, and the multiple photographing devices include at least two photographing devices arranged at different positions.
  • the target image includes the image of the target object in the image of the object;
  • the movement track of the target object is determined according to the target image.
  • the processor when it calls the program code, it also performs the following operation: displaying the movement track on an interactive interface.
  • the processor when the processor determines the target image from the images stored in the storage device, the processor includes performing the following operations:
  • the reference image of the target object is matched with the image of the object stored in the storage device to determine the target image from the images stored in the storage device.
  • the processor when the processor matches the reference image of the target object with the image of the object stored in the storage device to determine the target image from the images stored in the storage device, the processor includes performing the following operations:
  • the image corresponding to the matching coefficient in the stored images is determined as the target image.
  • processor when the processor calls the program code, it also performs the following operations:
  • the index information of the target object input by the user through the interactive interface is acquired, and the reference image of the target object is determined according to the index information of the target object.
  • the index information includes identity information of the target object
  • the processor obtains the index information of the target object input by the user through the interactive interface, and determines the reference image of the target object according to the index information of the target object , Including performing the following operations:
  • the image corresponding to the identity information is determined as the reference image of the target object.
  • the index information includes an image
  • the processor when the processor obtains the index information of the target object input by the user through the interactive interface, and determines the reference image of the target object according to the index information of the target object, the processor includes executing the following operating:
  • the image of the target object input by the user through the interactive interface is acquired, and the input image is determined as the reference image of the target object.
  • the processor determining the movement track of the target object according to the target image includes performing the following operations:
  • the movement track of the target object According to the position information of the target object, the movement track of the target object.
  • the processor determines the position of the target object according to the installation position of the photographing device that photographs the target image, it includes performing the following operations:
  • the location of the target object is determined according to the installation position and the shooting direction of the shooting device that shoots the target image, and the location of the target object in the shooting screen when the shooting device shoots the target image.
  • processor when the processor calls the program code, it also performs the following operations:
  • the image of the object is stored in the storage device.
  • the processor when determining the image of the object from the output image according to the object recognition algorithm, the processor further performs the following operations:
  • the image of the object is determined according to the bounding box.
  • the processor when determining the image of the object according to the bounding box, the processor further performs the following operations:
  • the remaining images in the target image after deletion are determined as the images of the target object.
  • the object is a pedestrian, and when the processor calls the program code, it also performs the following operations:
  • the image including the front face of the target object is determined from the target image, and the image including the front face of the target object is displayed.
  • processor when the processor calls the program code, it also performs the following operations:
  • second prompt information is displayed, where the second prompt information is used to indicate that the target row does not belong to the preset organization.
  • the target object is a pedestrian.
  • a storage medium on which program instructions are stored, and when the program instructions are run by a computer or a processor, they are used to execute the target object in the embodiment of the present invention.
  • the corresponding steps of the method for determining the movement trajectory are used to implement the corresponding modules in the device for determining the movement trajectory of the target object according to the embodiment of the present invention.
  • the storage medium may include, for example, a memory card of a smart phone, a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disk read-only memory (CD-ROM), USB memory, or any combination of the above storage media.
  • the computer-readable storage medium may be any combination of one or more computer-readable storage media.
  • one computer-readable storage medium contains computer-readable program code for randomly generating a sequence of action instructions
  • another computer may
  • the read storage medium contains computer-readable program codes for determining the movement trajectory of the target object.
  • each functional module of the device for determining the movement track of the target object according to the embodiment of the present invention can be realized, and/or can execute the operation of the target object according to the embodiment of the present invention.
  • the modules in the device for determining the movement trajectory of the target object according to the embodiment of the present invention can be implemented by computer program instructions stored in the memory by the processor of the electronic device that is determined according to the movement trajectory of the target object in the embodiment of the present invention, or
  • the computer instructions that can be stored in the computer-readable storage medium of the computer program product according to the embodiment of the present invention are implemented when the computer is running.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

本发明提供了一种目标对象的移动轨迹确定方法和装置，所述方法包括：从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；根据所述目标图像确定目标对象的移动轨迹。根据本发明提供的方法和装置，将多个拍摄装置拍摄到的目标对象的图像关联起来，通过多个拍摄装置拍摄到的目标对象的图像确定目标对象的运动轨迹，大大地减少了监控的工作量，提高了准确性。

Description

一种目标对象的移动轨迹确定方法和装置 技术领域
本发明涉及图像处理技术领域，尤其涉及一种目标对象的移动轨迹确定方法和装置。
背景技术
拍摄装置广泛地应用于监控领域。监控人员通过分析拍摄装置拍摄到的图像来对目标对象进行监控。然而，在监控领域，监控人员通过逐个分析拍摄到目标对象的拍摄装置输出的图像来实现对目标对象的移动轨迹分析，这样方式效率低下，工作强度大，准确性低。
发明内容
本发明实施例提供一种目标对象的移动轨迹确定方法和装置，以通过多个拍摄装置拍摄的图像确定目标对象的运动轨迹。
本发明实施例的第一方面，提供了一种目标对象的移动轨迹确定方法，包括：
从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
根据所述目标图像确定目标对象的移动轨迹。
本发明实施例的第二方面，提供一种目标对象的移动轨迹确定装置，包括：
存储器和处理器；
所述存储器用于存储程序代码；
所述处理器，调用所述程序代码，当程序代码被执行时，用于执行以下操作：
从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
根据所述目标图像确定目标对象的移动轨迹。
本发明实施例提供的目标对象的移动轨迹确定方法和装置，将多个拍摄装置拍摄到的目标对象的图像关联起来，通过多个拍摄装置拍摄到的目标对象的图像确定目标对象的运动轨迹，大大地减少了监控的工作量，提高了准确性。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种目标对象的移动轨迹确定方法的示意性流程图；
图2是本发明实施例提供的确定目标对象的移动轨迹的场景示意图；
图3是本发明实施例提供的显示目标对象的移动轨迹的示意图；
图4是本发明实施例提供的确定目标对象的位置的示意图；
图5是本发明实施例提供的确定目标对象相对于拍摄装置的方位的示意图；
图6是本发明实施例提供的删除边界框内的背景图像的示意图；
图7是本发明实施例提供的一种目标对象的移动轨迹确定装置的结构图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
下面结合附图，对本发明的一些实施方式作详细说明。在不冲突的情况下，下述的实施例及实施例中的特征可以相互组合。
参见图1，图1示出了根据本发明实施例的一种目标对象的移动轨迹确定方法的示意性流程图，其中，所述方法可以由目标对象的移动轨迹确定装置执行。如图1所示，本发明实施例提供一种目标对象的移动轨迹确定方法，所述方法100包括:
在步骤S110，从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
具体地,监控区域201(例如商场、办公楼等)可以设置多个拍摄装置201(例如包括至少两个拍摄装置)，其中，每一个拍摄装置可以设置在监控区域201的不同位置，另外，每一个拍摄装置201都可以对一个拍摄区域(例如如图所述的扇形区域)进行拍摄以获取图像。所述图像可以存储在存储装置中。在某些实施例中，服务器可以获取所述多个拍摄装置201拍摄获取的图像，并将所述图像存储在存储装置中。可选地，所述存储装置可以是服务器的本地存储装置，也可以是与所述服务器通信连接的存储装置。
对象203在监控区域201中移动时，对象203会被多个拍摄装置201拍摄到，即拍摄装置201会输出对象的图像，即，这样存储装置201会存储对象的图像。所述对象203可以为任何可以自主移动或者接受外力移动的物体，例如，所述对象可以为行人、动物、机器人等，在这里以对象为行人来进行示意性说明。对象包括目标对象，目标对象为对象中一个感兴趣或者指定的对象。存储装置存储目标对象的图像，即目标图像。目标对象的移动轨迹确定装置可以从存储装置存储的图像中确定目标对象的图像，即从存储装置存储的图像中筛选得到目标图像。
在步骤S120，根据所述目标图像确定目标对象的移动轨迹。
具体地，在确定所述目标图像之后，目标图像与所述拍摄装置201相关联，拍摄装置201又设置在监控区域201的不同位置，则目标图像可以用于指示目标对象在监控区域201中出现的位置。目标对象的移动轨迹确定装置可以根据所述目标图像确定目标对象在所述监控区域201中的移动轨迹204。
可选地，目标对象的移动轨迹确定装置可以为与所述存储装置通信连接装置，例如电脑等。可选地，所述目标对象的移动轨迹确定装置可以为所述服务器。
根据本发明实施例的目标对象的移动轨迹确定方法，将多个拍摄装置拍摄到的目标对象的图像关联起来，通过多个拍摄装置拍摄到的目标对象的图像确定目标对象的运动轨迹，大大地减少了监控的工作量，提高了准确性。
可选地，根据本发明实施例，所述方法还包括：在交互界面上显示所述移动轨迹。
具体地，目标对象的移动轨迹确定装置可以在交互界面上显示所述移动轨迹。其中，所述交互界面可以为显示屏。为了满足不同的用户需求，所述移动轨迹的显示可以通过轨迹线条图的形式显示，例如以显示如图2所示的移动轨迹204线条图的方式来显示；也可以通过文字的形式，例如如图3所示的方式来显示，如按照时间来显示目标对象到达的位置的形式显示所述移动轨迹。可以理解的是，本领域技术人员还可以采用其它任意可以显示时间及其对应的所述目标对象的位置的方式来显示所述移动轨迹，在此不做限制。
可选地，所述从存储装置存储的图像中确定目标图像，包括：将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像。
具体地，为了从存储装置存储的图像中确定目标图像，所述目标对象的移动轨迹确定装置可以获取目标对象的参考图像，然后将所述目标对象的参考图像与所述存储装置存储的图像进行匹配运算以从所述存储的图像中确定目标图像，进一步地，将所述目标对象的参考图像与所述存储装置存储的对象的图像进行匹配运算以从所述存储的对象的图像中确定目标图像。其中，所述目标对象的参考图像可以存储目标对象的移动轨迹确定装置的本地存储装置中，目标对象的参考图像也可以存储在与所述目标对象的移动轨迹确定装置通信连接的存储装置中，所述目标对象的移动轨迹确定装置可以从所述存储装置获取所述目标对象的参考图像。
可选地，所述将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像，包括：将目标对象的参考图像与所述存储装置存储的图像进行匹配以确定所述参考图像与所述存储的图像之间的匹配系数；当所述匹配系数满足预设条件时，将存储的图像中与所述匹配系数对应的图像确定为目标图像。
具体地，所述目标对象的移动轨迹确定装置可以将目标对象的参考图像与所述存储的图像进行匹配，通过所述匹配可以确定匹配系数，其中，所述匹配系数可以是表征所述存储的图像与所述参考图像之间的相似程度的参数。当所 述匹配系数满足预设条件时，确定所述存储的图像为所述目标图像，当所述匹配系数不满足预设条件时，确定所述存储的图像不是所述目标图像。
可选地，所述匹配系数包括存储的图像和参考图像之间的欧氏距离。其中，所述欧氏距离越小则表明所述存储的图像与参考图像中的相似程度越高，当存储的图像和参考图像之间的欧氏距离小于预设的欧式几何距离阈值时则表示存储的图像为目标图像。当存储的图像和参考图像之间的欧氏距离大于或等于预设的欧式几何距离阈值，确定则表示存储的图像不是所述目标图像。
具体地，目标对象的移动轨迹确定装置可以通过训练好的深度学习模型对存储的图像和参考图像提取特征，其中，所述特征为特征向量，这样可以得到存储的图像的特征X为(x 1，x 2，...x n)和参考图像的特征Y为(y 1，y 2，...y n)，则存储的图像和参考图像之间的欧氏距离
Figure PCTCN2019076557-appb-000001
可选地，所述提取特征可以采用诸如LBP(局部二值模式)、HoG(方向梯度直方图)、PCA(主成分分析法)或者神经网络等各种适当的特征提取方法来进行特征提取并生成所述多个特征向量。可选的，当所述对象为行人时，对于所述目标对象的参考图像和/或在存储装置存储的对象的图像中的每一帧图像中的对象的人脸均采用相同的特征提取方法生成特征向量。此处，可以将该人脸图像区域作为一个整体进行特征提取，也可以在该人脸图像区域的不同子图像区域中分别进行特征提取。
可选地，所述参考图像的特征向量与所述存储装置存储的图像的特征向量可以是经过归一化后的向量，即将向量标准化，使向量的模为1，这样有利于提高运算效率，便于快速检测到目标图像。
可选地，根据本发明实施例，所述方法还包括：获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像。
具体地，当用户向对某个特定的目标对象的移动轨迹进行查看时，用户可以在交互界面上输入这个目标对象的索引信息，其中，所述索引信息可以任何能够索引到目标对象的信息。所述目标对象的索引信息与所述目标对象的参考图像具有对应关系，目标对象的移动轨迹确定装置可以根据所述目标对象的索引信息确定所述目标对象的参考图像。
可选地，所述索引信息可以是目标对象的身份信息、生物特征信息、图像 中的至少一种。其中，所述身份信息包括工号、身份证号码、名字、保险账号等中的至少一种；所述生物特征信息可以包括指纹、声纹、虹膜等中的至少一种，另外，所述作为索引信息的目标对象的图像可以包括如前所述目标对象的参考图像。
可选地，所述索引信息包括目标对象的身份信息，所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像，包括：获取用户通过交互界面输入的目标对象的身份信息；根据所述身份信息获取与所述身份信息对应的图像；将所述身份信息对应的图像确定为目标对象的参考图像。
可选地，所述索引信息包括图像，所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像，包括：获取用户通过交互界面输入的目标对象的图像，将所述输入的图像确定为目标对象的参考图像。
需要说明的是，当所述索引信息包括图像时，所述图像可以是所述多个拍摄装置中至少一个拍摄装置输出的图像，也可以为通过其他方式获取的目标对象的图像。
可选地，在一个实施例中，所述目标对象的移动轨迹确定装置控制交互界面显示由多个拍摄装置拍摄显示的图像，根据接收到的用户输入的选择操作从所述显示的图像确定用户选中的图像，将所述用户选中的图像确定为目标对象的参考图像。通过这种方式，用户可以将拍摄装置输出的包括目标对象的图像选中并且确定为目标对象的参考图像。
可选地，所述根据所述目标图像确定目标对象的移动轨迹，包括：根据所述目标图像确定拍摄所述目标图像的拍摄装置；根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置信息；根据所述目标对象的位置信息目标对象的移动轨迹。
具体地，如前所述，至少两个拍摄装置安装在监控区域的不同位置，输出所述目标图像的拍摄装置的安装位置可以反映目标对象的位置信息。由于目标图像与拍摄所述目标图像的拍摄装置是一一对应的，目标对象的移动轨迹确定装置可以根据所述目标图像和所述对应关系确定拍摄所述目标图像的拍摄装置，并获取所述拍摄装置的安装位置，进而可以根据所述拍摄装置的安装位置 确定所述目标对象的位置信息。在确定所述目标对象的位置信息之后，即可以根据所述目标对象的位置信息确定目标对象的移动轨迹。在一些实施例中，在精度要求不高时，可以将所述目标图像的拍摄装置的安装位置确定为所述目标对象的位置。
可选地，所述根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置，包括：根据所述拍摄所述目标图像的拍摄装置的安装位置、拍摄方向和所述拍摄装置在拍摄所述目标图像时目标对象在拍摄画面中的位置确定目标对象的位置。
具体地，参见图4，拍摄装置401拍摄输出目标对象402的目标图像，目标对象402与拍摄装置之间的距离d可以根据所述目标图像确定，例如，目标图像中目标对象对应的图像区域(可能就是目标图像)的大小可以用于确定所述距离d。当所述图像区域较大时，说明目标对象402与拍摄装置之间的距离d较近，当所述图像区域较小时，说明目标对象402与拍摄装置之间的距离d较远。
另外，目标对象的移动轨迹确定装置可以根据拍摄装置的拍摄方向和所述拍摄装置在拍摄所述目标图像时目标对象在拍摄画面中的位置确定目标对象相对于所述拍摄装置的方位。请参见图5，目标对象在拍摄画面501中的位置为o 1，可以确定所述位置o 1与拍摄画面501的中心o之间的偏差，所述偏差可以包括水平方向上的偏差x和竖直方向上的偏差y。所述偏差可以反映目标对象偏离拍摄装置的拍摄方向的角度，根据所述偏差可以确定目标对象偏离拍摄装置的拍摄方向的角度，并进而根据所述拍摄装置的拍摄方向和所述角度确定目标对象相对于所述拍摄装置的方位。其中，所述方位可以至少包括俯仰方位和偏航方位。
进一步地，目标对象的移动轨迹确定装置可以根据所述距离d和所述方位确定目标对象相对于所述拍摄装置的位置，进而可以根据所述目标对象相对于所述拍摄装置的位置和所述拍摄装置的安装位置确定目标对象的位置。
可选地，所述方法还包括：获取所述多个拍摄装置输出的图像；根据对象识别算法从所述输出的图像中确定对象的图像；将所述对象的图像存储到所述存储装置中。
具体地，目标对象的移动轨迹确定装置可以获取所述拍摄装置输出的图 像。目标对象的移动轨迹确定装置可以针对获取到的每一帧图像或者按照预设的间隔周期性地对所述图像运行对象识别算法，通过这种方式从所述图像确定对象的图像。
可选地，所述根据对象识别算法从所述输出的图像中确定对象的图像，包括：根据对象识别算法确定对象在所述输出的图像中的边界框；根据所述边界框确定对象的图像。
具体地，参见图6，当拍摄装置输出的图601中包括对象时，通过所述识别算法可以从所述图像中确定对象在图像中的边界框602，例如，通过神经网络从所述图像中确定对象在图像中的边界框。进一步地，根据所述边界框确定对象的图像，在某些实施例中，可以将所述边界框内的图像确定为对象的图像。
进一步可选地，所述根据所述边界框确定对象的图像，包括：从边界框内的图像中识别背景的图像，删除所述背景的图像；将所述边界框内剩余图像确定为对象的图像。
具体地，继续参考图6，为了提高对如前所述的匹配的准确度，同时也为了降低数据存储量，目标对象的移动轨迹确定装置可以识别所述边界框602中的背景图像603，可以将背景的图像603中删除。这样，完成删除之后，可以将边界框602中剩下的图像604确定为对象的图像。
可选地，所述对象为行人，所述方法还包括：从所述目标图像中确定包括目标对象正脸的图像，显示所述包括目标对象正脸的图像。
具体地，为了准备对目标行人的身份进行识别，目标对象的移动轨迹确定装置从所述目标图像中确定包括目标对象正脸的图像，并控制交互界面显示所述目标对象正脸的图像。进一步地，还可以从所述目标对象正脸的图像确定正脸的图像区域，所述所述图像区域内的图像进行放大，然后控制所述交互界面显示放大后的图像以便更加清晰地识别目标行人的脸部。
可选地，可以通过人脸质量评估算法来从所述目标图像中确定包括目标对象正脸的图像，其中，人脸质量评估算法可以基于神经网络模型获取所述行人序列中的人脸图像的模糊程度和人脸的角度，并设置不同的权重，用来评估所述人脸图片质量。
其中，所述人脸图片的模糊程度可以是所述人脸图片和所述人脸图片的模糊图像的邻域梯度变化。清晰图像的纹理锐利，图像邻域变化显著；模糊图 像的邻域变化较小。在一个实施例中，将所述人脸图片通过预设的低通滤波器，得到所述人脸图片的模糊图像，计算所述模糊图像的邻域梯度变化；通过比较所述选取的人脸图片和所述模糊图像的邻域梯度变化，得到所述人脸图片的模糊程度。
人脸的角度可以通过深度学习基于卷积神经网络结构进行计算得到人脸的角度，或通过自适应随机森林算法分别对人脸的角度进行回归估计。
可选地，所述方法还包括：根据所述包括目标对象正脸的图像进行识别以确定所述目标对象是否属于预设的组织；当确定所述目标对象属于预设的组织时，显示第一提示信息，其中，所述第一提示信息用于指示目标对象属于预设的组织；当确定所述目标对象不属于预设的组织时，显示第二提示信息，其中，所述第二提示信息用于指示目标行不属于预设的组织。
具体地，目标对象的移动轨迹确定装置确定目标对象是否属于预设的组织，例如确定目标行人是否属于某个公司、部门、小组、业主群体等等，当确定所述目标对象属于预设的组织时，显示第一提示信息，当确定所述目标对象不属于预设的组织时，显示第二提示信息以实现警报的作用。
可选地，所述目标对象为行人。基于正脸图像进行人脸识别以确认所述目标对象的身份信息，进一步提高了行人识别的准确度，提高了识别效率。
参见图7，图7示出了根据本发明实施例的一种目标对象的移动轨迹确定装置的示意性框图。如图7所示，本发明实施例提供一种目标对象的移动轨迹确定装置，所述装置700包括:
存储器701和处理器702；
所述存储器用于存储程序代码；
所述处理器702，调用所述程序代码，当程序代码被执行时，用于执行以下操作：
从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
根据所述目标图像确定目标对象的移动轨迹。
可选地，所述处理器调用所述程序代码时，还执行如下操作：在交互界面 上显示所述移动轨迹。
可选地，所述处理器在从存储装置存储的图像中确定目标图像时，包括执行如下操作：
将目标对象的参考图像与所述存储装置存储的所述对象的图像进行匹配以从存储装置存储的图像中确定目标图像。
可选地，所述处理器在将目标对象的参考图像与所述存储装置存储的所述对象的图像进行匹配以从存储装置存储的图像中确定目标图像时，包括执行如下操作：
将目标对象的参考图像与所述存储装置存储的所述对象的图像进行匹配以确定所述参考图像与所述存储的图像之间的匹配系数；
当所述匹配系数满足预设条件时，将存储的图像中与所述匹配系数对应的图像确定为目标图像。
可选地，所述处理器调用所述程序代码时，还执行如下操作：
获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像。
可选地，所述索引信息包括目标对象的身份信息，所述处理器在所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像时，包括执行如下操作：
获取用户通过交互界面输入的目标对象的身份信息；
根据所述身份信息获取与所述身份信息对应的图像；
将所述身份信息对应的图像确定为目标对象的参考图像。
可选地，所述索引信息包括图像，所述处理器在所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像时，包括执行如下操作：
获取用户通过交互界面输入的目标对象的图像，将所述输入的图像确定为目标对象的参考图像。
可选地，所述处理器在根据所述目标图像确定目标对象的移动轨迹，包括执行如下操作：
根据所述目标图像确定拍摄所述目标图像的拍摄装置；
根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置；
根据所述目标对象的位置信息目标对象的移动轨迹。
可选地，所述处理器根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置时，包括执行如下操作：
根据所述拍摄所述目标图像的拍摄装置的安装位置、拍摄方向和所述拍摄装置在拍摄所述目标图像时目标对象在拍摄画面中的位置确定目标对象的位置。
可选地，所述处理器调用所述程序代码时，还执行如下操作：
获取所述多个拍摄装置输出的图像；
根据对象识别算法从所述输出的图像中确定对象的图像；
将所述对象的图像存储到所述存储装置中。
可选地，所述处理器根据对象识别算法从所述输出的图像中确定对象的图像时，还执行如下操作：
根据对象识别算法确定对象在所述输出的图像中的边界框；
根据所述边界框确定对象的图像。
可选地，所述处理器根据所述边界框确定对象的图像时，还执行如下操作：
从目标图像中识别背景的图像，删除所述背景的图像，其中，所述目标图像为所述输出的图像中的边界框内的图像；
将删除后目标图像中的剩余图像确定为目标对象的图像。
可选地，所述对象为行人，所述处理器调用所述程序代码时，还执行如下操作：
从所述目标图像中确定包括目标对象正脸的图像，显示所述包括目标对象正脸的图像。
可选地，所述处理器调用所述程序代码时，还执行如下操作：
根据所述包括目标对象正脸的图像进行识别以确定所述目标对象是否属于预设的组织；
当所述目标对象属于预设的组织时，显示第一提示信息，其中，所述第一提示信息用于指示目标对象属于预设的组织；
当所述目标对象不于预设的组织则，显示第二提示信息，其中，所述第二提示信息用于指示目标行不属于预设的组织。
可选地，所述目标对象为行人。
此外，根据本发明实施例，还提供了一种存储介质，在所述存储介质上存储了程序指令，在所述程序指令被计算机或处理器运行时用于执行本发明实施例的目标对象的移动轨迹确定方法的相应步骤，并且用于实现根据本发明实施例的目标对象的移动轨迹确定装置中的相应模块。所述存储介质例如可以包括智能电话的存储卡、平板电脑的存储部件、个人计算机的硬盘、只读存储器(ROM)、可擦除可编程只读存储器(EPROM)、便携式紧致盘只读存储器(CD-ROM)、USB存储器、或者上述存储介质的任意组合。所述计算机可读存储介质可以是一个或多个计算机可读存储介质的任意组合，例如一个计算机可读存储介质包含用于随机地生成动作指令序列的计算机可读的程序代码，另一个计算机可读存储介质包含用于进行目标对象的移动轨迹确定的计算机可读的程序代码。
在一个实施例中，所述计算机程序指令在被计算机运行时可以实现根据本发明实施例的目标对象的移动轨迹确定装置的各个功能模块，并且/或者可以执行根据本发明实施例的目标对象的移动轨迹确定方法。
根据本发明实施例的目标对象的移动轨迹确定装置中的各模块可以通过根据本发明实施例的目标对象的移动轨迹确定的电子设备的处理器运行在存储器中存储的计算机程序指令来实现，或者可以在根据本发明实施例的计算机程序产品的计算机可读存储介质中存储的计算机指令被计算机运行时实现。
尽管这里已经参考附图描述了示例实施例，应理解上述示例实施例仅仅是示例性的，并且不意图将本发明的范围限制于此。本领域普通技术人员可以在其中进行各种改变和修改，而不偏离本发明的范围和精神。所有这些改变和修改意在被包括在所附权利要求所要求的本发明的范围之内。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
本领域的技术人员可以理解，除了特征之间相互排斥之外，可以采用任何组合对本说明书(包括伴随的权利要求、摘要和附图)中公开的所有特征以及如此公开的任何方法或者设备的所有过程或单元进行组合。除非另外明确陈述，本说明书(包括伴随的权利要求、摘要和附图)中公开的每个特征可以由 提供相同、等同或相似目的的替代特征来代替。
此外，本领域的技术人员能够理解，尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征，但是不同实施例的特征的组合意味着处于本发明的范围之内并且形成不同的实施例。例如，在权利要求书中，所要求保护的实施例的任意之一都可以以任意的组合方式来使用。
应该注意的是上述实施例对本发明进行说明而不是对本发明进行限制，并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的元件或步骤。位于元件之前的单词“一”或“一个”不排除存在多个这样的元件。本发明可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。
以上所述，仅为本发明的具体实施方式或对具体实施方式的说明，本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。本发明的保护范围应以权利要求的保护范围为准。

Claims (30)

  1. 一种目标对象的移动轨迹确定方法，其特征在于，包括：
    从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
    根据所述目标图像确定所述目标对象的移动轨迹。
  2. 根据权利要求1所述的方法，其特征在于，所述方法还包括：在交互界面上显示所述移动轨迹。
  3. 根据权利要求1或2所述的方法，其特征在于，所述从存储装置存储的图像中确定目标图像，包括：
    将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像。
  4. 根据权利要求3所述的方法，其特征在于，所述将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像，包括：
    将目标对象的参考图像与所述存储装置存储的图像进行匹配以确定所述参考图像与所述存储的图像之间的匹配系数；
    当所述匹配系数满足预设条件时，将存储的图像中与所述匹配系数对应的图像确定为目标图像。
  5. 根据权利要求3或4所述的方法，其特征在于，所述方法还包括：
    获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像。
  6. 根据权利要求5所述的方法，其特征在于，所述索引信息包括目标对象的身份信息，所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像，包括：
    获取用户通过交互界面输入的目标对象的身份信息；
    根据所述身份信息获取与所述身份信息对应的图像；
    将所述身份信息对应的图像确定为目标对象的参考图像。
  7. 根据权利要求5所述的方法，其特征在于，所述索引信息包括图像，所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像，包括：
    获取用户通过交互界面输入的目标对象的图像，将所述输入的图像确定为目标对象的参考图像。
  8. 根据权利要求1所述的方法，其特征在于，所述根据所述目标图像确定目标对象的移动轨迹，包括：
    根据所述目标图像确定拍摄所述目标图像的拍摄装置；
    根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置；
    根据所述目标对象的位置信息目标对象的移动轨迹。
  9. 根据权利要求8所述的方法，其特征在于，所述根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置，包括：
    根据所述拍摄所述目标图像的拍摄装置的安装位置、拍摄方向和所述拍摄装置在拍摄所述目标图像时目标对象在拍摄画面中的位置确定目标对象的位置。
  10. 根据权利要求1-9任一项所述的方法，其特征在于，所述方法还包括：
    获取所述多个拍摄装置输出的图像；
    根据对象识别算法从所述输出的图像中确定对象的图像；
    将所述对象的图像存储到所述存储装置中。
  11. 根据权利要求10所述的方法，其特征在于，所述根据对象识别算法从所述输出的图像中确定对象的图像，包括：
    根据对象识别算法确定对象在所述输出的图像中的边界框；
    根据所述边界框确定对象的图像。
  12. 根据权利要求11所述的方法，其特者在于，所述根据所述边界框确定对象的图像，包括：
    从边界框内的图像中识别背景的图像，删除所述背景的图像；
    将所述边界框内剩余图像确定为对象的图像。
  13. 根据权利要求1-12任一项所述的方法，其特征在于，所述对象为行人，所述方法还包括：
    从所述目标图像中确定包括目标对象正脸的图像，显示所述包括目标对象正脸的图像。
  14. 根据权利要求13所述的方法，其特征在于，所述方法还包括：
    根据所述包括目标对象正脸的图像进行识别以确定所述目标对象是否属于预设的组织；
    当是时，显示第一提示信息，其中，所述第一提示信息用于指示目标对象属于预设的组织；
    否则，显示第二提示信息，其中，所述第二提示信息用于指示目标行不属于预设的组织。
  15. 根据权利要求1-12任一项所述的方法，其特征在于，所述目标对象为行人。
  16. 一种地目标对象的移动轨迹确定装置，其特征在于，所述装置包括:
    存储器和处理器；
    所述存储器用于存储程序代码；
    所述处理器，调用所述程序代码，当程序代码被执行时，用于执行以下操作：
    从存储装置存储的图像中确定目标图像，其中，所述存储装置存储的图像包括多个拍摄装置拍摄的对象的图像，所述多个拍摄装置至少包括两个设置在 不同位置的拍摄装置，所述目标图像包括对象的图像中目标对象的图像；
    根据所述目标图像确定所述目标对象的移动轨迹。
  17. 根据权利要求16所述的装置，其特征在于，所述处理器调用所述程序代码时，还执行如下操作：在交互界面上显示所述移动轨迹。
  18. 根据权利要求16或17所述的装置，其特征在于，所述处理器在从存储装置存储的图像中确定目标图像时，包括执行如下操作：
    将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像。
  19. 根据权利要求18所述的装置，其特征在于，所述处理器在将目标对象的参考图像与所述存储装置存储的图像进行匹配以从存储装置存储的图像中确定目标图像时，包括执行如下操作：
    将目标对象的参考图像与所述存储装置存储的图像进行匹配以确定所述参考图像与所述存储的图像之间的匹配系数；
    当所述匹配系数满足预设条件时，将存储的图像中与所述匹配系数对应的图像确定为目标图像。
  20. 根据权利要求18或19所述的装置，其特征在于，所述处理器调用所述程序代码时，还执行如下操作：
    获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像。
  21. 根据权利要求20所述的装置，其特征在于，所述索引信息包括目标对象的身份信息，所述处理器在所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像时，包括执行如下操作：
    获取用户通过交互界面输入的目标对象的身份信息；
    根据所述身份信息获取与所述身份信息对应的图像；
    将所述身份信息对应的图像确定为目标对象的参考图像。
  22. 根据权利要求20所述的装置，其特征在于，所述索引信息包括图像，所述处理器在所述获取用户通过交互界面输入的目标对象的索引信息，根据所述目标对象的索引信息确定目标对象的参考图像时，包括执行如下操作：
    获取用户通过交互界面输入的目标对象的图像，将所述输入的图像确定为目标对象的参考图像。
  23. 根据权利要求16所述的装置，其特征在于，所述处理器在根据所述目标图像确定目标对象的移动轨迹，包括执行如下操作：
    根据所述目标图像确定拍摄所述目标图像的拍摄装置；
    根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置；
    根据所述目标对象的位置信息目标对象的移动轨迹。
  24. 根据权利要求23所述的装置，其特征在于，所述处理器根据拍摄所述目标图像的拍摄装置的安装位置确定目标对象的位置时，包括执行如下操作：
    根据所述拍摄所述目标图像的拍摄装置的安装位置、拍摄方向和所述拍摄装置在拍摄所述目标图像时目标对象在拍摄画面中的位置确定目标对象的位置。
  25. 根据权利要求16-24任一项所述的装置，其特征在于，所述处理器调用所述程序代码时，还执行如下操作：
    获取所述多个拍摄装置输出的图像；
    根据对象识别算法从所述输出的图像中确定对象的图像；
    将所述对象的图像存储到所述存储装置中。
  26. 根据权利要求25所述的装置，其特征在于，所述处理器根据对象识别算法从所述输出的图像中确定对象的图像时，还执行如下操作：
    根据对象识别算法确定对象在所述输出的图像中的边界框；
    根据所述边界框确定对象的图像。
  27. 根据权利要求26所述的装置，其特者在于，所述处理器根据所述边界框确定对象的图像时，还执行如下操作：
    从边界框内的图像中识别背景的图像，删除所述背景的图像；
    将所述边界框内剩余图像确定为对象的图像。
  28. 根据权利要求16-27任一项所述的装置，其特征在于，所述对象为行人，所述处理器调用所述程序代码时，还执行如下操作：
    从所述目标图像中确定包括目标对象正脸的图像，显示所述包括目标对象正脸的图像。
  29. 根据权利要求28所述的装置，其特征在于，所述处理器调用所述程序代码时，还执行如下操作：
    根据所述包括目标对象正脸的图像进行识别以确定所述目标对象是否属于预设的组织；
    当所述目标对象属于预设的组织时，显示第一提示信息，其中，所述第一提示信息用于指示目标对象属于预设的组织；
    当所述目标对象不于预设的组织则，显示第二提示信息，其中，所述第二提示信息用于指示目标行不属于预设的组织。
  30. 根据权利要求16-29任一项所述的装置，其特征在于，所述目标对象为行人。
PCT/CN2019/076557 2019-02-28 2019-02-28 一种目标对象的移动轨迹确定方法和装置 Ceased WO2020172870A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2019/076557 WO2020172870A1 (zh) 2019-02-28 2019-02-28 一种目标对象的移动轨迹确定方法和装置
CN201980005227.6A CN111263955A (zh) 2019-02-28 2019-02-28 一种目标对象的移动轨迹确定方法和装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/076557 WO2020172870A1 (zh) 2019-02-28 2019-02-28 一种目标对象的移动轨迹确定方法和装置

Publications (1)

Publication Number Publication Date
WO2020172870A1 true WO2020172870A1 (zh) 2020-09-03

Family

ID=70953441

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/076557 Ceased WO2020172870A1 (zh) 2019-02-28 2019-02-28 一种目标对象的移动轨迹确定方法和装置

Country Status (2)

Country Link
CN (1) CN111263955A (zh)
WO (1) WO2020172870A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112044075A (zh) * 2020-09-18 2020-12-08 网易（杭州）网络有限公司 运动轨迹的生成方法和装置
CN114171211A (zh) * 2021-12-14 2022-03-11 平安国际智慧城市科技股份有限公司 一种轨迹追踪方法、装置、计算机设备及存储介质
CN116883003A (zh) * 2023-07-10 2023-10-13 国家电网有限公司客户服务中心 基于生物探针技术的移动端支付购电反欺诈方法、系统
CN117241133A (zh) * 2023-11-13 2023-12-15 武汉益模科技股份有限公司 基于非固定位置的多工序同时作业的视觉报工方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111970434A (zh) * 2020-07-22 2020-11-20 吉林省智擎工业软件研究院有限公司 多摄像机多目标的运动员跟踪拍摄视频生成系统及方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160125249A1 (en) * 2014-10-30 2016-05-05 Toyota Motor Engineering & Manufacturing North America, Inc. Blur object tracker using group lasso method and apparatus
CN105898205A (zh) * 2015-01-04 2016-08-24 伊姆西公司 用于由多个摄像头监视目标对象的方法和装置
CN106127137A (zh) * 2016-06-21 2016-11-16 长安大学 一种基于3d轨迹分析的目标检测识别算法
CN108769598A (zh) * 2018-06-08 2018-11-06 复旦大学 基于行人重识别的跨摄像头视频浓缩方法
CN108875548A (zh) * 2018-04-18 2018-11-23 科大讯飞股份有限公司 人物轨迹生成方法及装置、存储介质、电子设备
CN108876821A (zh) * 2018-07-05 2018-11-23 北京云视万维科技有限公司 跨镜头多目标跟踪方法及系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107358146B (zh) * 2017-05-22 2018-05-22 深圳云天励飞技术有限公司 视频处理方法、装置及存储介质
CN108256443A (zh) * 2017-12-28 2018-07-06 深圳英飞拓科技股份有限公司 一种人员定位方法、系统及终端设备
CN109166261B (zh) * 2018-10-11 2022-06-07 平安科技（深圳）有限公司 基于图像识别的图像处理方法、装置、设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160125249A1 (en) * 2014-10-30 2016-05-05 Toyota Motor Engineering & Manufacturing North America, Inc. Blur object tracker using group lasso method and apparatus
CN105898205A (zh) * 2015-01-04 2016-08-24 伊姆西公司 用于由多个摄像头监视目标对象的方法和装置
CN106127137A (zh) * 2016-06-21 2016-11-16 长安大学 一种基于3d轨迹分析的目标检测识别算法
CN108875548A (zh) * 2018-04-18 2018-11-23 科大讯飞股份有限公司 人物轨迹生成方法及装置、存储介质、电子设备
CN108769598A (zh) * 2018-06-08 2018-11-06 复旦大学 基于行人重识别的跨摄像头视频浓缩方法
CN108876821A (zh) * 2018-07-05 2018-11-23 北京云视万维科技有限公司 跨镜头多目标跟踪方法及系统

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112044075A (zh) * 2020-09-18 2020-12-08 网易（杭州）网络有限公司 运动轨迹的生成方法和装置
CN114171211A (zh) * 2021-12-14 2022-03-11 平安国际智慧城市科技股份有限公司 一种轨迹追踪方法、装置、计算机设备及存储介质
CN116883003A (zh) * 2023-07-10 2023-10-13 国家电网有限公司客户服务中心 基于生物探针技术的移动端支付购电反欺诈方法、系统
CN117241133A (zh) * 2023-11-13 2023-12-15 武汉益模科技股份有限公司 基于非固定位置的多工序同时作业的视觉报工方法及系统
CN117241133B (zh) * 2023-11-13 2024-02-06 武汉益模科技股份有限公司 基于非固定位置的多工序同时作业的视觉报工方法及系统

Also Published As

Publication number Publication date
CN111263955A (zh) 2020-06-09

Similar Documents

Publication Publication Date Title
CN105518709B (zh) 用于识别人脸的方法、系统和计算机程序产品
US20170262472A1 (en) Systems and methods for recognition of faces e.g. from mobile-device-generated images of faces
JP4241763B2 (ja) 人物認識装置及びその方法
WO2019033574A1 (zh) 电子装置、动态视频人脸识别的方法、系统及存储介质
WO2019109526A1 (zh) 人脸图像的年龄识别方法、装置及存储介质
WO2019033569A1 (zh) 眼球动作分析方法、装置及存储介质
US11176679B2 (en) Person segmentations for background replacements
JP7419080B2 (ja) コンピュータシステムおよびプログラム
WO2021012370A1 (zh) 瞳孔半径的检测方法、装置、计算机设备和存储介质
CN110472460B (zh) 人脸图像处理方法及装置
CN111263955A (zh) 一种目标对象的移动轨迹确定方法和装置
CN111931594A (zh) 人脸识别活体检测方法、装置、计算机设备和存储介质
US10650234B2 (en) Eyeball movement capturing method and device, and storage medium
JP5361524B2 (ja) パターン認識システム及びパターン認識方法
WO2019033568A1 (zh) 嘴唇动作捕捉方法、装置及存储介质
WO2019033570A1 (zh) 嘴唇动作分析方法、装置及存储介质
JP2013504114A (ja) 目状態検出装置及び方法
JP2016031679A (ja) オブジェクト識別装置、オブジェクト識別方法及びプログラム
CN108171138B (zh) 一种生物特征信息获取方法和装置
KR20190093799A (ko) Cctv를 통한 실시간 실종자 얼굴 인식 시스템 및 그 방법
CN114511893B (zh) 一种卷积神经网络的训练方法、人脸识别方法及装置
CN113436735A (zh) 基于人脸结构度量的体重指数预测方法、设备和存储介质
CN112347988A (zh) 口罩识别模型的训练方法、装置、计算机设备和可读存储介质
WO2020115910A1 (ja) 情報処理システム、情報処理装置、情報処理方法、およびプログラム
CN108875549B (zh) 图像识别方法、装置、系统及计算机存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19916757

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19916757

Country of ref document: EP

Kind code of ref document: A1