WO2024002340A1 - 对待测物中的目标对象进行追踪的方法及系统 - Google Patents

对待测物中的目标对象进行追踪的方法及系统 Download PDF

Info

Publication number
WO2024002340A1
WO2024002340A1 PCT/CN2023/104626 CN2023104626W WO2024002340A1 WO 2024002340 A1 WO2024002340 A1 WO 2024002340A1 CN 2023104626 W CN2023104626 W CN 2023104626W WO 2024002340 A1 WO2024002340 A1 WO 2024002340A1
Authority
WO
WIPO (PCT)
Prior art keywords
viewing device
target object
mobile viewing
dimensional
target
Prior art date
Application number
PCT/CN2023/104626
Other languages
English (en)
French (fr)
Inventor
季峥
卢杰梅
刘磊
马媛
喻卫丰
宗春光
Original Assignee
同方威视技术股份有限公司
清华大学
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 同方威视技术股份有限公司, 清华大学 filed Critical 同方威视技术股份有限公司
Publication of WO2024002340A1 publication Critical patent/WO2024002340A1/zh

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/66—Tracking systems using electromagnetic waves other than radio waves
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
    • G01N23/02—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material
    • G01N23/04—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and forming images of the material
    • G01N23/046—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and forming images of the material using tomography, e.g. computed tomography [CT]
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/88—Lidar systems specially adapted for specific applications
    • G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging
    • G01S17/894—3D imaging with simultaneous measurement of time-of-flight at a 2D array of receiver pixels, e.g. time-of-flight cameras or flash lidar

Definitions

  • At least one embodiment of the present disclosure relates to a method for tracking a target object in an object to be measured, and in particular, to a method and system for tracking a target object in an object to be measured.
  • Fluoroscopic imaging is an indispensable means in the field of security inspection, and digital radiography (Digital-Radiography, DR) and computed tomography (Computed Tomography, CT) are both advantageous technologies for realizing material discrimination.
  • the approximate position of the contraband at a certain viewing angle or several viewing angles can only be determined based on perspective imaging, and then the search is performed manually.
  • the current inspection method cannot understand the changes in the relative position of the inspector and the contraband in real time during the manual tracking process. , and it is impossible to provide more effective guidance to inspection personnel.
  • embodiments of the present disclosure provide a method and system for tracking a target object in an object to be measured, so as to quickly and accurately find the target object.
  • a method for tracking a target object in an object to be measured including: identifying the target object in a perspective image and determining a three-dimensional model of the target object in the object to be measured. First position information; using a mobile viewing device to construct a three-dimensional appearance image of the object to be measured, and determining second position information of the mobile viewing device relative to the three-dimensional appearance image; according to the first position information and the second position information, Get the The three-dimensional coordinates of the target object relative to the mobile viewing device; and associating the three-dimensional coordinates to the mobile viewing device, and viewing the relative position of the mobile viewing device to the target object in real time through the mobile viewing device to guide the movement of the mobile viewing device The user approaches the target object.
  • the method further includes obtaining first information and a perspective image of the object to be tested.
  • using a mobile viewing device to construct a three-dimensional shape image of the object under test includes: identifying the first information; and constructing a three-dimensional image of the object under test corresponding to the first information. 3D shape image.
  • obtaining a fluoroscopic image of the object under test includes scanning the object under test using X-rays to obtain the fluoroscopic image.
  • obtaining a fluoroscopic image of the object under test includes scanning the object under test through X-ray computed tomography to obtain the three-dimensional fluoroscopic image.
  • identifying the target object in the perspective image and determining the first position information of the three-dimensional model of the target object in the object includes: identifying and identifying the target object in the perspective image. Mark the target object in the object to be tested to obtain a marked target in one viewing angle; add a new marking target by adding a viewing angle and supplementary marking to obtain at least one marked target in another viewing angle; and according to the viewing angle and the described Mark the target and determine the first position information of the target object within the three-dimensional model of the space defined by the object to be measured.
  • identifying the target object in the perspective image and determining the first position information of the three-dimensional model of the target object in the object to be measured includes:
  • the first position information within the three-dimensional model of the space defined by the object.
  • constructing a three-dimensional shape image of the object under test includes: using a mobile viewing device to obtain features of multiple outer surfaces of the object under test; and based on a plurality of the outer surfaces Features construct a three-dimensional shape image of the object under test.
  • obtaining the three-dimensional coordinates of the target object relative to the mobile viewing device according to the first position information and the second position information includes: taking a point on the object to be measured as the origin Constructing a three-dimensional coordinate system of the three-dimensional appearance image; generating a first coordinate point of the first position information in the three-dimensional coordinate system; generating a second coordinate point of the second position information in the three-dimensional coordinate system; and obtaining the The three-dimensional coordinates of the first coordinate point relative to the second coordinate point in the three-dimensional coordinate system.
  • the three-dimensional coordinates are associated with a mobile viewing device, and the relative position of the mobile viewing device to the target object is viewed in real time through the mobile viewing device to guide the movement of the mobile viewing device.
  • the user approaching the target object includes: associating the three-dimensional coordinates with the mobile viewing device; setting the perspective of the mobile viewing device as the initial tracking position, and setting the direction from the initial tracking position to the target object as a first tracking direction; and the mobile viewing device moves along the first tracking direction, so that the distance between the tracking position of the mobile viewing device and the target object is reduced, and is reduced to a preset value Stop moving.
  • the three-dimensional coordinates are associated with a mobile viewing device, and the relative position of the mobile viewing device to the target object is viewed in real time through the mobile viewing device to guide the user of the mobile viewing device
  • Approaching the target object further includes: moving the mobile viewing device along a second tracking direction different from the first tracking direction, and the distance between the moved tracking position of the mobile viewing device and the target object increases.
  • the moving direction of the mobile viewing device is adjusted to the first tracking direction and moves along the first tracking direction, so that the distance between the moved tracking position of the mobile viewing device and the target object is reduced.
  • a system for tracking a target object in an object under test including: a processor, including: a recognition module configured to identify and mark the object under test in a perspective image; The target object in the object to determine the first position information of the three-dimensional model of the target object in the object to be tested; and a mobile viewing device configured to obtain a three-dimensional appearance image of the object to be tested; wherein the movement The viewing device is further configured to determine second position information of the mobile viewing device relative to the three-dimensional appearance image; and obtain the three-dimensional coordinates of the target object relative to the mobile viewing device based on the first position information and the second position information; The three-dimensional coordinates are associated with the mobile viewing device, and the relative position of the mobile viewing device to the target object is viewed in real time through the mobile viewing device, so as to guide the user of the mobile viewing device to approach the target object.
  • the system further includes a scanning device adapted to scan the object to be tested to obtain a perspective image of the object to be tested.
  • the scanning device includes an X-ray scanning imaging device, adapted to acquire a perspective image of at least one viewing angle of the object to be measured.
  • the scanning device includes an X-ray tomography device, adapted to acquire a three-dimensional perspective image of the object to be measured.
  • the system further includes a collection device, adapted to collect the first information of the object to be tested.
  • the recognition module is further configured to: identify the target object in the object to be measured in the perspective image, and mark the target object to obtain at least two viewing angles. the marking target; and determining the first position information of the target object within the three-dimensional model of the space defined by the object to be measured according to the viewing angle and the marking target.
  • the mobile viewing device includes a lidar module configured to acquire characteristics of multiple external surfaces of the object to be measured, and construct the object to be measured based on the characteristics of the multiple external surfaces. Three-dimensional image of the object.
  • the mobile viewing device is further configured to: construct a three-dimensional coordinate system of the three-dimensional shape image using a point on the object to be measured as an origin; and generate the third coordinate system in the three-dimensional coordinate system.
  • a first coordinate point of position information generating a second coordinate point of the second position information in a three-dimensional coordinate system; and obtaining a three-dimensional coordinate point of the first coordinate point relative to the second coordinate point in the three-dimensional coordinate system. coordinate.
  • the mobile viewing device is configured to move in a second tracking direction that is different from the first tracking direction, and the tracking position after movement of the mobile viewing device corresponds to the target object.
  • the distance between them increases, stop moving; and adjust the moving direction of the mobile viewing device to the first tracking direction, and move along the first tracking direction, so that the tracking position of the mobile viewing device after movement is consistent with the The distance between target objects decreases, stopping when it reaches a preset value.
  • the target object in the perspective image is identified to determine the position information of the three-dimensional model of the target object in the object to be measured.
  • the perspective image and the three-dimensional shape image of the measured object are correlated to register the position of the target object in the three-dimensional space.
  • the position of the target object in the three-dimensional space is associated with the mobile viewing device, so that the user can understand the relative position of the mobile viewing device and the target object through the mobile viewing device and guide the user to get closer.
  • Target objects are used to improve the convenience and accuracy of users' tracking of target objects.
  • Figure 1 schematically shows a flow chart of a method for tracking a target object in an object to be measured according to an exemplary embodiment of the present disclosure
  • Figure 2 schematically shows a flow chart of a method for tracking a target object in an object to be measured according to another exemplary embodiment of the present disclosure
  • FIG. 3 schematically shows a flow chart of an implementation method of step S200 in Figure 2;
  • FIG. 4 schematically shows a flow chart of another implementation method of step S200 in Figure 2;
  • Figure 6 schematically shows a partial display effect of a three-dimensional shape image according to an embodiment of the present disclosure
  • FIG. 7 schematically illustrates a block diagram of a system for tracking a target object in an object to be measured according to an exemplary embodiment of the present disclosure
  • Figure 8 schematically shows a block diagram of the system of an exemplary embodiment shown in Figure 7;
  • Figure 9 schematically shows a block diagram of the system of another illustrative embodiment shown in Figure 7;
  • Figure 10 schematically shows a structural diagram of a scanning device according to an exemplary embodiment of the present disclosure
  • Figure 11 schematically shows a structural diagram of a scanning device according to another exemplary embodiment of the present disclosure.
  • FIG. 12 schematically shows a structural diagram of a three-dimensional shape image acquisition device according to an exemplary embodiment.
  • a tracking method for a target object in an object to be measured is provided.
  • FIG. 1 schematically shows a flow chart of a method for tracking a target object in an object to be measured according to an exemplary embodiment of the present disclosure.
  • a method for tracking a target object in an object to be measured including identifying the target object in a perspective image and determining where the target object is in the object to be measured.
  • the target object may be displayed on the mobile viewing device 2 while the user is viewing the relative position of the mobile viewing device 2 to the target object in real time through the mobile viewing device 2 .
  • the target object may be concretely displayed in the mobile viewing device 2 .
  • the target object may be displayed in a visual form in a display component (eg, a display screen) of the mobile viewing device 2 as a point in a three-dimensional coordinate system or an area based on the outline of the target object.
  • a point on the target object may be used as the point of the three-dimensional coordinate system displayed in the mobile viewing device 2 .
  • the acquired perspective images include perspective images of containers (including ISO air containers), cargo trucks, and other objects to be tested.
  • the tracking method further includes obtaining first information and a perspective image of the object to be measured.
  • FIG. 2 schematically shows a flowchart of a method for tracking a target object in an object to be measured according to another exemplary embodiment of the present disclosure.
  • the tracking method includes steps S100 to S600.
  • the collection device 3 is used to collect the first information of the object to be tested (such as a container, a cargo truck), the scanning device 4 is used to obtain a perspective image of the object to be tested, and the first information and the object to be tested corresponding to the first information are The perspective image of the object is associated (such as establishing a mapping relationship on the data) and stored in the database for retrieval.
  • the first information is represented by information used to identify the object to be tested, including but not limited to at least one of the label number of the container, the license plate number of the freight truck, and the information code corresponding to the object to be tested.
  • step S200 the processor 1 identifies the target object in the perspective image to determine the first position information of the three-dimensional model of the target object in the object to be measured, and associates the first position information with the perspective image and stores it in in the database to be retrieved.
  • the three-dimensional model in the object to be tested includes but is not limited to the first three-dimensional model constructed by taking the length, width and height directions of the inner wall of the accommodation space defined by the object to be tested as the X-axis, Y-axis and Z-axis of the three-dimensional model respectively.
  • the first position information includes but is not limited to coordinate points in the first three-dimensional model established based on the object to be measured.
  • target objects include but are not limited to explosives, flammable materials, drugs, firearms, and prohibited items such as controlled knives. It should be noted that in step S200, if the target object (contraband) is not recognized in the perspective image of the object to be tested, the object to be tested can be directly released without performing subsequent steps.
  • the mobile viewing device 2 is used to identify the second information, construct a three-dimensional shape image of the object to be measured, and determine the second position information of the mobile viewing device 2 relative to the three-dimensional shape image.
  • the second information is characterized by information used to identify the object to be tested, including but not limited to at least one of the label number of the container, the license plate number of the freight truck, and the information code corresponding to the object to be tested.
  • the second position information includes but is not limited to coordinate points in the second three-dimensional model established based on the three-dimensional shape image.
  • step S400 the first position information corresponding to the first information matching the second information stored in the database and/or the perspective image containing the first position information is retrieved according to the second information.
  • step S500 the three-dimensional coordinates of the target object relative to the mobile viewing device are obtained. Since the position of the accommodation space defined by the object under test is relatively unique relative to the external shape of the object under test, the first three-dimensional model (including the first position information) constructed based on the accommodation space of the object under test is relatively unique. The position of the three-dimensional shape image of the measured object is also relatively unique. Moreover, the three-dimensional shape image is constructed by the mobile viewing device 2. The relative position of the mobile viewing device 2 and the three-dimensional shape image is known, and the mobile viewing device 2 can be used as another point in the three-dimensional coordinate system.
  • the corresponding first position information and the three-dimensional shape image can be associated (including but not limited to data association or image association established by using at least one of pointer and mark information) to change the coordinates of the target object. reflected in the second three-dimensional model. That is to say, the relative position between two points (the target object and the mobile viewing device 2) in the three-dimensional coordinate system can be known, and the relative position of the mobile viewing device 2 to the target object can be viewed in real time through the mobile viewing device 2. position to guide the user of the mobile viewing device 2 to approach the target object.
  • the three-dimensional coordinates of the target object relative to the mobile viewing device include directly related or indirectly related three-dimensional coordinates.
  • the directly related three-dimensional coordinates are represented by the three-dimensional coordinates of the target object relative to the mobile viewing device; the indirectly related three-dimensional coordinates are represented by the first three-dimensional coordinates and the third point of the target object relative to the target object and a third point other than the mobile viewing device.
  • the three-dimensional coordinates of the target object relative to the mobile viewing device can be calculated based on the first three-dimensional coordinates and the second three-dimensional coordinates.
  • the third point includes but is not limited to a point on the object to be measured, and the third point can be used as the origin of the three-dimensional coordinate system for establishing the second three-dimensional model.
  • step S600 associate the three-dimensional coordinates to the mobile viewing device 2 (including but not limited to displaying the three-dimensional coordinates on the mobile viewing device 2 and constructing image information of the three-dimensional coordinate system to express the three-dimensional coordinates into the three-dimensional coordinate system) .
  • the vehicle is scanned by an X-ray imaging device to obtain a perspective image of the vehicle. If it is recognized that there are prohibited objects in the vehicle, and currently only one marked target is obtained from one perspective, in this case, you can set other perspectives and manually add new marked targets to ensure that the prohibited objects are detected from at least two perspectives. Mark to get at least two views of the marked target. Finally, according to the viewing angle and the marked target, the first position information of the three-dimensional model of the target object in the object to be measured is confirmed.
  • the geometric center of the target object is at the first position of the three-dimensional model in the object to be measured according to the viewing angle (that is, the incident direction of the X-ray). location information.
  • the first position information of the target object includes the first position information of the geometric center of the target object.
  • the perspective image of the object to be tested is acquired during the security inspection, and the first information is associated with the perspective image. Thereafter, the target object in the perspective image is identified. If the acquired perspective image shows that there is a target object (contraband), a three-dimensional appearance image of the object to be measured is further acquired. If there is no target object (prohibited object) in the obtained fluoroscopic image, the security inspection of the object to be tested is completed. In this way, there is no need to construct a three-dimensional shape image for each object to be measured, which can effectively reduce the amount of data processing.
  • it includes but is not limited to taking a point in the object to be measured (such as the midpoint of a certain side wall of the container, at least one of the front license plate or the rear license plate of the freight truck) as the origin, obtaining the relative position of the target object, The three-dimensional coordinates of the origin.
  • a point in the object to be measured such as the midpoint of a certain side wall of the container, at least one of the front license plate or the rear license plate of the freight truck
  • the relative position of the origin and the object to be measured is fixed, establishing a three-dimensional coordinate system based on the origin is conducive to calibrating other coordinate points in the three-dimensional coordinate system (including but not limited to target objects and/or mobile viewing). device).
  • the above-mentioned three-dimensional appearance image is established by the mobile viewing device 2.
  • the mobile viewing device 2 can also be regarded as a point in the three-dimensional coordinate system. Therefore, it can be known that the relative relationship between the mobile viewing device 2 and, for example, the above-mentioned origin point position, and then the relative position (three-dimensional coordinates) of the target object relative to the mobile viewing device 2 can be obtained.
  • obtaining a perspective image of the object to be tested includes, but is not limited to, obtaining a perspective image by scanning the object to be tested with X-rays.
  • FIG. 3 schematically shows a flow chart of an implementation method of step S200 in FIG. 2 .
  • an implementation method of step S200 includes the following sub-steps S210 to S240.
  • step S210 the identification module of the processor 1 identifies the target object in the object to be measured in the perspective image, and marks the target object. It should be noted here that the specific method for identifying the target object is not regarded as the key point of protection of the present disclosure. Any image identification method known in the art can be selected and applied, and will not be further elaborated.
  • step S220 the processor 1 determines whether the target object has markers for two or more viewing angles.
  • step S230 the position information of the target object in the three-dimensional model of the space defined by the object to be measured is determined based on the viewing angles and the marked targets.
  • step S240 a new marked target is added by changing the perspective and supplementary marking, to obtain at least one marked target from another perspective, and return to sub-step S220.
  • the perspective image is obtained based on one viewing angle, it only limits one plane in space. Therefore, other perspective images need to be obtained from another viewing angle, and then the perspective image obtained from this viewing angle is The target object is marked manually, thereby calibrating the position information of the target object in the three-dimensional model. If the perspective image is obtained based on two different perspectives (such as two orthogonal perspectives), a three-dimensional model can be formed in space.
  • obtaining a fluoroscopic image of the object to be tested includes, but is not limited to, obtaining a fluoroscopic image through X-ray computed tomography scanning of the object to be tested.
  • FIG. 4 schematically shows a flow chart of another implementation method of step S200 in FIG. 2 .
  • CT scan X-ray computed tomography
  • the CT scan used in this implementation method eliminates the need to acquire multiple perspective images from multiple viewing angle images and obtain marked targets from multiple viewing angles.
  • the following sub-steps S250 to S260 are specifically included.
  • step S260 the first position information of the target object in the three-dimensional tomographic image is determined based on the first three-dimensional model established by the three-dimensional tomographic image.
  • X-ray computed tomography itself has a three-dimensional imaging function, so there is no need to By determining the number of viewing angles, the first position information of the target object can be obtained.
  • FIG. 5 schematically shows a flow chart of an implementation method of step S300 in FIG. 2 .
  • step S300 In an implementation method of step S300, the following sub-steps S310 to S330 are included.
  • step S320 features of multiple outer surfaces of the object to be measured can be obtained by moving the viewing device 2.
  • step S330 a three-dimensional shape image of the object to be measured is constructed based on the characteristics of the multiple outer surfaces.
  • steps S320 and S330 include using the lidar module of the mobile viewing device 2 to collect original point cloud data of the outer surface of the object to be measured, and then registering, splicing, and removing the original point cloud data.
  • Noise simplification and other processes are performed to obtain the solid point cloud data of the object to be tested, and then modeling is performed through the solid point cloud data (including but not limited to at least one of geometry creation, plane creation and mapping) to generate a three-dimensional representation of the object to be tested.
  • Shape image including but not limited to at least one of geometry creation, plane creation and mapping
  • step S340 includes selecting a point on the object to be measured as an origin, establishing a second three-dimensional model based on the origin and the three-dimensional shape image of the object to be measured, and using the mobile viewing device 2 as the third A point in the two-dimensional and three-dimensional models is reflected in the three-dimensional coordinate system established based on the second three-dimensional model.
  • constructing a three-dimensional shape image of the object to be tested includes using the three-dimensional shape image acquisition device 5 to acquire characteristics of multiple outer surfaces of opposite sides and the top of the object to be tested. According to the common features in two adjacent appearance images, multiple appearance images are spliced into a three-dimensional appearance image of the object to be measured.
  • a photographing device 52 such as a camera, is installed on the mast 51.
  • three photographing devices 52 are used to capture at least the appearance images of the opposite sides and top of the vehicle.
  • Types of vehicles may include, but are not limited to, large trucks, light trucks, sport utility vehicles (SUVs), sedans, etc.
  • the three-dimensional appearance image acquisition device 5 is independent of the mobile viewing device 2 and can be arranged side by side where the object to be tested is subject to security inspection, so as to acquire the three-dimensional appearance image while the object to be tested is undergoing security inspection. In this way, the user does not need to actively construct a three-dimensional shape image, which can effectively reduce the user's workload.
  • FIG. 6 schematically shows a partial display effect of a three-dimensional shape image according to an embodiment of the present disclosure.
  • the vehicle appearance panoramic splicing device 53 Splice multiple appearance images according to common features in two adjacent appearance images captured by the same shooting device 52, thereby splicing multiple appearance images into a panoramic image of the vehicle, where the panoramic image may include The front surface, the rear surface, the left surface (shown in (a) of FIG. 6 ), the right surface (shown in (b) of FIG. 6 ), the upper surface (shown in (c) of FIG. 6 ), and the lower surface of the vehicle.
  • the position of the three-dimensional model relative to the object under test is unique.
  • the three-dimensional model is associated with the three-dimensional shape image of the object to be measured, and a three-dimensional coordinate system is constructed based on a point on the object to be measured (such as the midpoint of the front license plate of a freight truck) as the origin.
  • the position information (such as coordinate points) of the target object in the three-dimensional model is mapped to the three-dimensional coordinate system to obtain the three-dimensional coordinates of the target object in the three-dimensional coordinate system.
  • the perspective image and the three-dimensional appearance image can be spliced to obtain a stitched image.
  • the steps of obtaining the combined image include: obtaining a plurality of first identification features of the object under test based on the perspective image; obtaining a plurality of second identification features of the object under test based on the three-dimensional shape image; obtaining based on the first identification feature and the second identification feature.
  • Combine the images use a point on the object to be measured as the origin to construct a three-dimensional coordinate system for the combined image; obtain the three-dimensional coordinates of the target object within the three-dimensional coordinate system.
  • the first identification feature includes features of an edge portion of a perspective image
  • the second identification feature includes features of an edge portion of a three-dimensional contour image
  • 5 feature points on the edge contour of the vehicle are obtained based on the perspective image as 5 first identification features, and then 5 feature points on the edge contour feature of the vehicle are obtained based on the three-dimensional image as 5 second identification features, align the 5 first identification features and the 5 second identification features to obtain multiple stitched images, and select the stitched image with the highest alignment accuracy between the first identification features and the second identification features .
  • the perspective image and the three-dimensional contour image are spliced into a stitched image.
  • the combined image is expressed in a three-dimensional space, and the target object and the mobile viewing device 2 are both visualized in the three-dimensional coordinate system (for example, the position of the user's perspective in the mobile viewing device 2 is used as the second coordinate system A coordinate point, the mobile viewing device 2 displays the relative position of the target object and the mobile viewing device in the three-dimensional coordinate system).
  • the distance between the mobile viewing device 2 and the target object can be concretely reflected, so that the user's tracking process is more intuitive.
  • the three-dimensional coordinates are associated with the mobile viewing device 2 , and the relative position of the mobile viewing device 2 to the target object is detected in real time by the mobile viewing device 2 to guide the use of the mobile viewing device 2
  • the user's approach to the target object includes: associating the three-dimensional coordinates with the mobile viewing device 2; setting the perspective of the mobile viewing device 2 as the initial tracking position, and setting the direction from the initial tracking position to the target object as the first tracking direction; moving the viewing device 2 moves along the first tracking direction so that the distance between the moved tracking position of the mobile viewing device 2 and the target object decreases, and stops moving when it decreases to a preset value.
  • the preset value may be set to half the minimum distance in a certain direction of the object to be measured, for example, half the length of the container in the width direction.
  • the mobile viewing device has positioned the target object relatively accurately and cannot or does not need to move further.
  • associating the three-dimensional coordinates to the mobile viewing device 2 includes displaying the three-dimensional coordinate system in the mobile viewing device 2 in video and/or image form.
  • the mobile viewing device 2 is further configured to operate on the displayed three-dimensional coordinate system, and the operation is characterized by being able to enlarge, reduce or rotate a local or all area of the three-dimensional coordinate system (including using a certain coordinate point in the three-dimensional coordinate system to Rotation in three-dimensional space and/or rotation around an axis of the three-dimensional coordinate system).
  • the three-dimensional coordinates are associated with the mobile viewing device 2, and the relative position of the mobile viewing device 2 to the target object is detected in real time by the mobile viewing device 2, so as to guide the user of the mobile viewing device 2 to approach the target.
  • the object also includes: when the mobile viewing device 2 moves along a second tracking direction different from the first tracking direction, and the distance between the tracking position of the mobile viewing device 2 and the target object increases after movement, stop the movement and adjust the mobile viewing
  • the moving direction of the device 2 is to the first tracking direction and moves along the first tracking direction, so that the distance between the tracking position of the mobile viewing device 2 and the target object is reduced.
  • the mobile viewing device 2 may vibrate, sound a ringtone, or appear a text warning, etc. to prompt the user to adjust the movement of the mobile viewing device 2 direction.
  • the mobile viewing device 2 can display three-dimensional coordinates including the target object, and the user can determine the mobile viewing device based on the three-dimensional coordinates. 2 to the tracking direction of the target object, and then the user can hold the mobile viewing device 2 and move toward the target object of the target object; when the mobile viewing device 2 moves along the first tracking direction, the mobile viewing device 2 moves The distance between the tracking position and the target object decreases, stopping when it reaches a preset value. Thereafter, usual manual inspection can be used to confirm the specific location of the target object and confirm whether it is a prohibited item.
  • the user can hold the mobile viewing device 2 and move it and display the relative three-dimensional coordinates of the target object on the mobile viewing device 2 in real time, and can basically determine the specific position of the target object in the object to be measured, thereby achieving accurate tracking of the target object. and positioning, which greatly shortens the time of manual tracking, improves the readability of images, and makes it easier to find suspects quickly and accurately.
  • a system for tracking a target object in an object to be measured is also provided.
  • FIG. 7 schematically shows a block diagram of a system for tracking a target object in an object under test according to an exemplary embodiment of the present disclosure.
  • a system for tracking a target object in an object to be measured includes: a processor 1 and a mobile viewing device 2 .
  • the processor 1 includes a recognition module configured to identify and mark the target object in the object under test in the perspective image to determine the first position information of the target object.
  • the mobile viewing device 2 is configured to obtain a three-dimensional shape image of the object to be measured, determine second position information of the mobile viewing device 2 relative to the three-dimensional shape image, and obtain the relative position of the target object relative to the mobile viewing device based on the first position information and the second position information. 2, and associate the three-dimensional coordinates to the mobile viewing device 2, and detect the relative position of the mobile viewing device 2 to the target object in real time through the mobile viewing device 2, so as to guide the user of the mobile viewing device 2 to approach the target object.
  • the mobile viewing device 2 includes a lidar module.
  • the lidar module is configured to collect the original point cloud data of the outer surface of the object to be measured, and then perform registration, splicing, denoising and simplification on the original point cloud data to obtain the entity point cloud data of the object to be measured, and then use the entity points to Model the cloud data (including but not limited to at least one of geometry creation, plane creation, and mapping) to generate a three-dimensional shape image of the object to be measured.
  • lidar modules include, but are not limited to, lidar scanners.
  • the obtained perspective images include perspective images of containers (including ISO air containers), cargo trucks and other objects to be tested.
  • the fluoroscopic image obtained here may be a two-dimensional fluoroscopic image obtained by X-ray scanning or a three-dimensional fluoroscopic image obtained by CT scanning. It is difficult to determine the target object based on a single two-dimensional perspective image obtained by X-ray scanning at one viewing angle. Therefore, it is necessary to perform X-ray scanning based on multiple viewing angles to obtain the corresponding two-dimensional perspective image at each viewing angle, and then based on multiple two-dimensional perspective images Determine target audience. Multiple here generally refers to at least two.
  • the mobile viewing device 2 may display the relative position of the target object and the mobile viewing device 2 . In this way, the user can hold the mobile viewing device 2 to track the target object on the three-dimensional image, which improves the readability of the image and facilitates finding the suspect more quickly and accurately.
  • a system for tracking a target object in an object to be measured may optionally include a scanning device 4 configured to acquire, for example, containers (including ISO air containers), cargo trucks, and the like.
  • a scanning device 4 configured to acquire, for example, containers (including ISO air containers), cargo trucks, and the like.
  • Perspective image of the object under test may be obtained, and then the perspective image of the object to be tested may be stored in a database, and the perspective image of the object to be tested stored in the database may be directly retrieved during this use.
  • FIG. 8 schematically shows a block diagram of the system of an exemplary embodiment shown in FIG. 7 .
  • the system for tracking a target object in an object to be measured also includes a scanning device 4 adapted to scan the object to be measured to obtain a perspective image of the object to be measured.
  • the scanning device 4 includes an X-ray scanning imaging device, adapted to acquire perspective images of at least two viewing angles of the object to be measured.
  • the mobile viewing device 2 may include a handheld terminal device, such as, but not limited to, a smartphone, a tablet, a laptop computer, a desktop computer, etc., or may be a variety of unmanned electronic devices. , including but not limited to drones, robots, etc.
  • the mobile viewing device 2 uses a smartphone, and the user can view the current tracking position of the user on the screen of the smartphone, and use this position as the initial tracking position. Users can slide the phone screen up, down, left, and right and manually click on the desired location to determine the tracking location. Users can also slide two fingers relative or away from each other on the screen to achieve local zooming, making it easier to view details. More specifically, the mobile viewing device 2 may be a smart terminal with a LiDAR scanner, such as an iPhone or iPad with LiDAR.
  • the system for tracking a target object in an object to be tested also includes a collection device 3 adapted to collect first information of the object to be tested. More specifically, the collection device 3 may be a camera, such as a visible light camera.
  • the camera identifies the front license plate number and/or the rear license plate number of the vehicle (first information).
  • the camera identifies the container number (first information) on the ISO air container.
  • the camera identifies the container number (first information) on the vehicle-mounted ISO aviation container.
  • the system for tracking the target object in the object to be measured also includes a verification module.
  • the verification module and the identification module are communicatively connected and configured to verify the first information of the object under test.
  • the verification module can be a camera integrated in the mobile viewing device 2 and an image recognition component (including image recognition software) connected to the camera, configured to collect at least a part of the object to be tested (the part containing the second information). , such as a license plate) image information, and recognize the collected image information to identify the second information, and match and verify it with the first information stored in the database.
  • the collection device 3 is suitable for collecting the first information of the object to be tested
  • the verification module is suitable for collecting the second information of the object to be tested, and comparing the second information with the first information to verify the first information. authenticating.
  • the system for tracking a target object in an object to be measured also includes a retrieval module in the processor 1 (not shown in the figure).
  • the retrieval module is configured to retrieve the first position information of the object to be tested and/or the perspective image containing the first position information corresponding to the first information from the database according to the first information of the object to be tested.
  • the object to be tested includes a container or a vehicle, and the first information of the object to be tested includes a container number or a vehicle license plate number.
  • the retrieval module may be a functional module that retrieves data from a database or a cloud platform that stores relevant information of the object under test (including the three-dimensional model and the position information within the three-dimensional model) through pipeline communication.
  • the first position information corresponding to the first information and/or the perspective image containing the first position information stored in the database is retrieved.
  • FIG. 10 schematically shows a structural diagram of a scanning device according to an exemplary embodiment of the present disclosure.
  • the scanning device 4 includes a first radiation source 40 that emits X-rays to the object to be measured, such as a container, at one viewing angle, and two sets of detector arrays installed on the gantry 54 (not shown in the figure). ) receives X-rays passing through the object to be measured, thereby obtaining a perspective image of the object to be measured at one viewing angle (height direction as shown in Figure 10).
  • the direction of the object under test can be changed (for example, the object under test is set to pass through the radiation position of the first radiation source 40 again along the length direction, as shown in Fig. (not shown), or adjust the angle at which the first radiation source 40 emits rays, or set up another set of first radiation sources 40 to obtain a perspective image at another viewing angle.
  • the recognition module is configured to: identify the target object in the object under test in the perspective image, and mark the target object to obtain the marked target in two or more viewing angles; and according to the viewing angle and the marked target, Determine the position information of the target object in the perspective image.
  • the recognition module is configured to: identify the target object in the object under test in the perspective image, and mark the target object to obtain a marked target under a viewing angle; and add new markings by adding new viewing angles. target, obtain the marked target under more than two viewing angles; and determine the first position information of the three-dimensional model of the target object in the object to be measured based on the viewing angle and the marked target.
  • the recognition module is configured to: identify the target object in the object under test in the perspective image, and mark the target object to obtain the marked target; and determine the position of the target object in the perspective image according to the viewing angle and the marked target. First location information.
  • the scanning device 4 includes an X-ray tomography device, which is suitable for acquiring a three-dimensional perspective image of an object to be measured.
  • the scanning device 4 is a CT scan-based X-ray imaging device.
  • CT scan is a scanning method that uses computer technology to reconstruct the tomographic image of the object being measured to obtain a three-dimensional tomographic image. Should scan The scanning method penetrates the measured object through a single axial plane of rays. According to the different absorption and transmittance of the rays of each part of the measured object, the computer performs three-dimensional reconstruction imaging based on the transmitted rays received by the detector. Based on this, CT The scanning fluoroscopic image acquisition device can acquire three-dimensional fluoroscopic images using only a set of ray sources and detectors.
  • FIG. 11 schematically shows a structural diagram of a scanning device according to another exemplary embodiment of the present disclosure.
  • the scanning device 4 includes a second radiation source 41 suitable for emitting X-rays, a corrector 42, a front collimator 43, an annular rotating frame 44, and a detector installed on the annular rotating frame 44.
  • the annular rotating frame 44 is a large rotatable annular ring.
  • the second radiation source 41 and the detector array 45 are both fixed on the annular rotating frame 44.
  • the annular rotating frame 44 Rotate and scan to obtain a three-dimensional perspective image of the container.
  • the recognition module is further configured to identify the target object in the object to be measured in the perspective image, and mark the target object to obtain the marked target in at least two viewing angles. According to the viewing angle and the marked target, the position information of the target object within the three-dimensional model of the space defined by the object to be measured is determined.
  • the recognition module is further configured to identify the target object in the object to be measured in the perspective image, and mark the target object. Get the marked target in one perspective, add new marked targets by changing the perspective, supplement the markers, and get the marked targets in at least two perspectives. According to the viewing angle and the marked target, the position information of the target object within the three-dimensional model of the space defined by the object to be measured is determined.
  • the verification module includes, but is not limited to, a camera integrated in the mobile viewing device 2 and an image recognition component (including image recognition software) communicated with the camera.
  • FIG. 9 schematically shows a block diagram of the system of another exemplary embodiment shown in FIG. 7 .
  • the system for tracking the target object in the object to be measured also includes a stitching module configured to construct a stitched image based on the perspective image and the three-dimensional shape image, and construct the three-dimensional coordinates of the stitched image in the stitched image. Tie.
  • the stitching module is further configured to acquire a plurality of first identification features of the object under test based on the perspective image and obtain a plurality of second identification features of the object under test based on the three-dimensional shape image. According to the first identification feature and the second identification feature, a stitched image is obtained.
  • the first identification feature includes features of the edge portion of the perspective image
  • the second identification feature includes features of the edge portion of the three-dimensional contour image
  • the three-dimensional appearance image acquisition device includes: a mast 51 , at least three shooting devices 52 and a vehicle appearance panoramic splicing device 53 .
  • the portal frame 51 includes at least two side columns 511 and/or at least one cross beam 512. Both ends of the cross beam 512 are connected to the side columns 511 respectively.
  • At least three photographing devices 52 are respectively disposed on two opposite side pillars 511 and cross beams 512.
  • the container vehicle drives through the mast 51, and the photographing devices 52 photograph multiple appearance images of the vehicle according to a preset frame rate.
  • the vehicle appearance panoramic splicing device 53 is communicatively connected to the shooting device 52 and is used to splice multiple appearance images into sub-images based on common features in two adjacent appearance images shot by each shooting device 52, and then splice into sub-images based on the multiple sub-images. A three-dimensional image of the vehicle under test.
  • the mobile viewing device 2 can prompt the user to change the moving direction. Specifically, if the mobile viewing device 2 moves along a second tracking direction that is different from the first tracking direction and the distance between the moved tracking position of the mobile viewing device 2 and the target object increases, then the movement is stopped; further, the movement is adjusted.
  • the moving direction of the viewing device 2 is to the modified first tracking direction, and moves along the modified first tracking direction, so that the distance between the moved tracking position of the mobile viewing device 2 and the target object is reduced. Stop movement when setting value.
  • the position of the target object in the three-dimensional coordinate system is fixed, and the position of the mobile viewing device changes according to the user's position. Based on the position change of the mobile viewing device, the user's coordinates can be updated in real time. , and reacquire the three-dimensional coordinates relative to the first coordinate point (target object) based on the updated second coordinate point to correct the position of the user and the target object.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Electromagnetism (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Health & Medical Sciences (AREA)
  • Radiology & Medical Imaging (AREA)
  • Pulmonology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Theoretical Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Image Analysis (AREA)

Abstract

一种对待测物中的目标对象进行追踪的方法及系统，对待测物中的目标对象进行追踪的方法包括对透视图像中的目标对象进行识别，并确定目标对象在待测物中的三维模型的第一位置信息；利用移动查看装置（2）构建待测物的三维外形图像，确定移动查看装置（2）相对于三维外形图像的第二位置信息；根据第一位置信息和第二位置信息，获取目标对象相对于移动查看装置（2）的三维坐标；以及将三维坐标关联至移动查看装置（2），通过移动查看装置（2）实时查看移动查看装置（2）至目标对象的相对位置，以引导移动查看装置（2）的使用者靠近目标对象。对待测物中的目标对象进行追踪的系统包括处理器（1）及移动查看装置（2），处理器（1）包括识别模块。

Description

对待测物中的目标对象进行追踪的方法及系统
本申请要求于2022年7月1日提交的、申请号为202210774556.0的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本公开的至少一种实施例涉及一种对待测物中的目标对象进行追踪的方法，特别是，涉及一种对待测物中的目标对象进行追踪的方法及系统。
背景技术
透视成像是安全检查领域必不可少的手段，数字X线摄影(Digital-Radiography，DR)、计算机断层扫描(Computed Tomography，CT)均是实现物质分辨的优势技术。
特别是，在例如适用于集装箱、货物卡车之类的大型设备进行安全检查期间，如果通过射线检查设备检测到集装箱或者货物卡车中存在例如爆炸物、毒品、枪支、管制刀具之类的违禁物的情况下，需要对违禁品进行具体进行查验和判定。现有的查验方法中，一般只能通过透视成像判断待测物中存在疑似的违禁品，具体的判定还需要在待测物中找到违禁品后进行人为确认。
目前，对于违禁品在待测物中所处的具体位置尚没有有效的定位方法，一般只能基于透视成像确定违禁品在某一视角或某几视角的大致方位，再通过人工方式进行寻找。然而，由于集装箱和货物卡车的容量很大，内部可能容纳有许多物品，在这种情况下，目前所采用的查验方法无法在人工追踪的过程中实时了解查验人员与违禁品的相对位置的变化，也无法对查验人员进行较为有效的引导。
发明内容
鉴于现有技术的上述或者其他方面的至少一种技术问题，本公开的实施例，提供一种对待测物中的目标对象进行追踪的方法及系统，以快速精准找到目标对象。
根据本公开的一个方面，提供了一种对待测物中的目标对象进行追踪的方法，包括：对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息；利用移动查看装置构建所述待测物的三维外形图像，确定所述移动查看装置相对于所述三维外形图像的第二位置信息；根据第一位置信息和第二位置信息，获取所 述目标对象相对于移动查看装置的三维坐标；以及将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象。
在本公开的一些实施例中，所述方法还包括获取待测物的第一信息及透视图像。
在本公开的一些实施例中，所述利用移动查看装置构建所述待测物的三维外形图像包括：识别所述第一信息；以及构建与所述第一信息对应的所述待测物的三维外形图像。
在本公开的一些实施例中，所述获取待测物的透视图像包括通过X射线扫描待测物、获取所述透视图像。
在本公开的一些实施例中，获取待测物的透视图像包括通过X射线计算机断层扫描待测物、获取三维的所述透视图像。
在本公开的一些实施例中，所述对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息包括：在所述透视图像中识别并标记所述待测物中的目标对象，得到一个视角下的标记目标；通过增加视角，补充标记添加新的标记目标，得到至少一个另一个视角下的标记目标；以及根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
在本公开的一些实施例中，所述对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息包括：
在所述透视图像中识别并标记所述待测物中的目标对象，得到至少两个视角下的标记目标；以及根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
在本公开的一些实施例中，所述构建所述待测物的三维外形图像包括：利用移动查看装置获取所述待测物的多个外表面的特征；以及根据多个所述外表面的特征构建所述待测物的三维外形图像。
在本公开的一些实施例中，所述根据第一位置信息和第二位置信息，获取所述目标对象相对于所述移动查看装置的三维坐标包括：以所述待测物上的一点作为原点构建所述三维外形图像的三维坐标系；在三维坐标系中生成所述第一位置信息的第一坐标点；在三维坐标系中生成所述第二位置信息的第二坐标点；以及获取所述第一坐标点相对于第二坐标点在所述三维坐标系中的三维坐标。
在本公开的一些实施例中，所述将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的 使用者靠近所述目标对象包括：将所述三维坐标关联至移动查看装置中；将所述移动查看装置的视角设为初始追踪位置，将所述初始追踪位置至所述目标对象的方向设为第一追踪方向；以及所述移动查看装置沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象之间的距离减小，在减小至预设值时停止移动。
在本公开的一些实施例中，所述将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象还包括：在所述移动查看装置沿与第一追踪方向不同的第二追踪方向移动，且所述移动查看装置移动后的追踪位置与所述目标对象之间的距离增大时，调整所述移动查看装置的移动方向至第一追踪方向，沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象的距离减小。
在本公开的另一个方面，提供了一种用于对待测物中的目标对象进行追踪的系统，包括：处理器，包括：识别模块，被配置成在透视图像中识别并标记所述待测物中的目标对象，以确定所述目标对象在待测物中的三维模型的第一位置信息；以及移动查看装置，被配置成获取所述待测物的三维外形图像；其中，所述移动查看装置还被配置成确定所述移动查看装置相对于所述三维外形图像的第二位置信息；根据第一位置信息和第二位置信息，获取所述目标对象相对于移动查看装置的三维坐标；将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象。
在本公开的一些实施例中，所述系统还包括扫描装置，适用于扫描所述待测物，以获取所述待测物的透视图像。
在本公开的一些实施例中，所述扫描装置包括X射线扫描成像设备，适用于获取所述待测物的至少一个视角的透视图像。
在本公开的一些实施例中，所述扫描装置包括X射线断层扫描设备，适用于获取所述待测物的三维的透视图像。
在本公开的一些实施例中，所述系统还包括采集装置，适用于采集所述待测物的第一信息。
在本公开的一些实施例中，所述移动查看装置还包括验证模块，被配置成验证所述待测物的第一信息。
在本公开的一些实施例中，所述处理器还包括调取模块，配置为从数据库中调取与第一信息对应的待测物中的目标对象的第一位置信息；所述待测物包括集装箱或车辆； 所述待测物的第一信息包括集装箱编号或车辆车牌号。
在本公开的一些实施例中，所述识别模块进一步被配置成：在所述透视图像中识别所述待测物中的目标对象，并对所述目标对象进行标记，得到一个视角下的标记目标；通过增加视角，补充标记添加新的标记目标，得到至少两个视角下的标记目标；以及根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
在本公开的一些实施例中，所述识别模块进一步被配置成：在所述透视图像中识别所述待测物中的目标对象，并对所述目标对象进行标记，得到至少两个视角下的标记目标；以及根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
在本公开的一些实施例中，所述移动查看装置包括激光雷达模块，被配置成获取所述待测物的多个外表面的特征，并根据多个外表面的所述特征构建所述待测物的三维外形图像。
在本公开的一些实施例中，所述移动查看装置进一步被配置成：以所述待测物上的一点作为原点构建所述三维外形图像的三维坐标系；在三维坐标系中生成所述第一位置信息的第一坐标点；在三维坐标系中生成所述第二位置信息的第二坐标点；以及获取所述第一坐标点相对于第二坐标点在所述三维坐标系中的三维坐标。
在本公开的一些实施例中，所述移动查看装置被配置成：在沿与第一追踪方向不同的第二追踪方向移动，且所述移动查看装置移动后的追踪位置与所述目标对象对应之间的距离增大时，停止移动；以及调整所述移动查看装置的移动方向至第一追踪方向，沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象之间的距离减小，在减小至预设值时停止移动。
根据本公开上述实施例的对待测物中的目标对象进行追踪的方法及系统，对透视图像中的目标对象进行识别，以确定目标对象在待测物中的三维模型的位置信息，将针对待测物的透视图像及三维外形图像进行关联，以对目标对象在三维空间内所处的位置配准。并且，基于配准后的信息将目标对象在三维空间内所处的位置关联至移动查看装置中，以使得使用者能通过移动查看装置了解移动查看装置与目标对象的相对位置并引导使用者靠近目标对象，用以提升使用者对目标对象进行追踪的便利性及准确性。
附图说明
通过以下参照附图对本公开实施例的描述，本公开的上述内容以及其他目的、特征和优点将更为清楚，在附图中：
图1示意性示出了根据本公开的一种示意性实施例的对待测物中的目标对象进行追踪的方法的流程图；
图2示意性示出了根据本公开的另一种示意性实施例的对待测物中的目标对象进行追踪的方法的流程图；
图3示意性示出了图2中步骤S200的一种实施方法的流程图；
图4示意性示出了图2中步骤S200的另一种实施方法的流程图；
图5示意性示出了图2中步骤S300的一种实施方法的流程图；
图6示意性示出了根据本公开实施例的三维外形图像的局部显示效果示意图；
图7示意性示出了根据本公开的一种示意性实施例的用于对待测物中的目标对象进行追踪的系统的方框图；
图8示意性示出了图7所示的一种示意性实施例的系统的方框图；
图9示意性示出了图7所示的另一种示意性实施例的系统的方框图；
图10示意性示出了根据本公开的一种示意性实施例中扫描装置的结构示意图；
图11示意性示出了根据本公开的另一种示意性实施例中扫描装置的结构示意图；以及
图12示意性示出了一种示意性实施例的三维外形图像获取装置的结构示意图。
具体实施方式
以下，将参照附图来描述本公开的实施例。但是应该理解，这些描述只是示意性的，而并非要限制本公开的范围。在下面的详细描述中，为便于解释，阐述了许多具体的细节以提供对本公开实施例的全面理解。然而，明显地，一个或多个实施例在没有这些具体细节的情况下也可以被实施。此外，在以下说明中，省略了对公知结构和技术的描述，以避免不必要地混淆本公开的概念。
在此使用的术语仅仅是为了描述具体实施例，而并非意在限制本公开。在此使用的术语“包括”、“包含”等表明了所述特征、步骤、操作和/或部件的存在，但是并不排除存在或添加一个或多个其他特征、步骤、操作或部件。
在此使用的所有术语(包括技术和科学术语)具有本领域技术人员通常所理解的含义，除非另外定义。应注意，这里使用的术语应解释为具有与本说明书的上下文相一致 的含义，而不应以理想化或过于刻板的方式来解释。
在使用类似于“A、B和C等中至少一个”这样的表述的情况下，一般来说应该按照本领域技术人员通常理解该表述的含义来予以解释(例如，“具有A、B和C中至少一个的系统”应包括但不限于单独具有A、单独具有B、单独具有C、具有A和B、具有A和C、具有B和C、和/或具有A、B、C的系统等)。
根据本公开的一种总体上的发明构思，提供一种对待测物中的目标对象的追踪方法。
图1示意性示出了根据本公开的一种示意性实施例的对待测物中的目标对象进行追踪的方法的流程图。
参考图1所示，根据本公开的一种示意性实施例，提供一种对待测物中的目标对象进行追踪的方法，包括对透视图像中的目标对象进行识别，并确定目标对象在待测物中的三维模型的第一位置信息；利用移动查看装置2构建待测物的三维外形图像，确定移动查看装置相对于三维外形图像的第二位置信息；根据第一位置信息和第二位置信息，获取目标对象相对于移动查看装置2的三维坐标；将三维坐标关联至移动查看装置2，通过移动查看装置2实时查看移动查看装置至目标对象的相对位置，以引导移动查看装置2的使用者靠近目标对象。
在本公开的一些实施例中，使用者在通过移动查看装置2实时查看移动查看装置2至目标对象的相对位置的状态下，目标对象可显示于移动查看装置2上。
在一个具体实施例中，目标对象可具象地显示于移动查看装置2中。例如，目标对象可以以可视的形式在移动查看装置2的显示部件(例如显示屏)中显示为三维坐标系中的一个点或基于目标对象的轮廓的区域。更具体地，可采用目标对象上的一点作为显示于移动查看装置2中的三维坐标系的点。例如，目标对象的几何中心、目标对象的轮廓上的点和/或目标对象内部的点。
在另一个具体实施例中，目标对象可以以相对于移动查看装置2的距离、方向和/或其他参数形式显示于移动查看装置2中。例如，以位于视角的某一侧及位于视角延伸方向的一侧的最短距离来显示目标对象。
在本公开的一些实施例中，已经获取的透视图像包括集装箱(包括ISO航空集装箱)、货物卡车之类的待测物的透视图像。
在本公开的另一些实施例中，追踪方法还包括获取待测物的第一信息及透视图像。
图2示意性示出了根据本公开的另一种示意性实施例的对待测物中的目标对象进行追踪的方法的流程图。
参考图2所示，在一些具体的实施例中，追踪方法包括步骤S100至S600。
在步骤S100中，利用采集装置3采集待测物(如集装箱、货物卡车)的第一信息，利用扫描装置4获取待测物的透视图像，将第一信息及与第一信息对应的待测物的透视图像进行关联(如建立数据上的映射关系)并存储于数据库中以待调取。其中，第一信息表征为用于识别待测物的信息，包括但不限于集装箱的标号、货运卡车的车牌号及与待测物对应的信息码中的至少一种。
在步骤S200中，通过处理器1对透视图像中的目标对象进行识别，以确定目标对象在待测物中的三维模型的第一位置信息，并将第一位置信息与透视图像关联并存储于数据库中以待调取。待测物中的三维模型包括但不限于以待测物所限定的容纳空间的内壁的长向、宽向及高向分别作为三维模型的X轴、Y轴及Z轴所构建的第一三维模型。第一位置信息包括但不限于依据待测物建立的第一三维模型内的坐标点。其中，目标对象包括但不限于爆炸物、易燃物、毒品、枪支、以及管制刀具之类的违禁物。应予说明，在步骤S200中，若在待测物的透视图像中没有识别到目标对象(违禁物)，则可直接放行该待测物而无需进行后续的步骤。
在步骤S300中，利用移动查看装置2识别第二信息，构建待测物的三维外形图像，并确定移动查看装置2相对于三维外形图像的第二位置信息。第二信息表征为用于识别待测物的信息，包括但不限于集装箱的标号、货运卡车的车牌号及与待测物对应的信息码中的至少一种。第二位置信息包括但不限于依据三维外形图像建立的第二三维模型内的坐标点。
在步骤S400中，依据第二信息调取数据库中所存储的与第二信息匹配的第一信息对应的第一位置信息和/或包含有第一位置信息的透视图像。
在步骤S500中，获取目标对象相对于移动查看装置的三维坐标。由于待测物所限定的容纳空间相对于待测物的外部形状的位置是相对唯一的，因此，依据待测物的容纳空间所构建的第一三维模型(包括第一位置信息)相对于待测物的三维外形图像的位置也是相对唯一的。而且，三维外形图像是通过移动查看装置2构建的，移动查看装置2和三维外形图像的相对位置是可知的，可将移动查看装置2作为三维坐标系中的另一点。因此，可将对应的第一位置信息及三维外形图像关联(包括但不限于利用指针及标示信息中的至少一种所建立的数据上的关联或图像上的关联)，以将目标对象的坐标体现在第二三维模型中。也就是说，三维坐标系中的两点(目标对象及移动查看装置2)之间的相对位置可知，可通过移动查看装置2实时查看移动查看装置2至目标对象的相对位 置，引导移动查看装置2的使用者靠近目标对象。目标对象相对于移动查看装置的三维坐标包括直接相关的或间接相关的三维坐标。直接相关的三维坐标表征为目标对象相对于移动查看装置的三维坐标；间接相关的三维坐标表征为目标对象相对于目标对象及移动查看装置之外的第三点的第一三维坐标及第三点相对于移动查看装置的第二三维坐标，基于第一三维坐标及第二三维坐标可通过计算得出目标对象相对于移动查看装置的三维坐标。第三点包括但不限于待测物上的一点，可将该第三点作为建立第二三维模型的三维坐标系的原点。
在步骤S600中，将三维坐标关联至移动查看装置2(包括但不限于将三维坐标在移动查看装置2上进行显示及、构建三维坐标系的图像信息以将三维坐标表达至三维坐标系中)。
在一个具体实施例中，通过X射线成像设备扫描车辆，获取车辆的透视图像。如识别到车辆中存在违禁物，并且当前只得到一个视角下的标记目标，这种情况下可以通过增加设置其他视角，手动补充标记添加新的标记目标，保证在至少两个视角下对违禁物进行标记，以得到至少两个视角下的标记目标。最后根据视角和标记目标，确认目标对象在待测物中的三维模型的第一位置信息。在此，例如在目标对象存在两个视角下的标记目标的情况下，根据视角(即X射线的入射方向)，能够确定该目标对象的几何中心在在待测物中的三维模型的第一位置信息。在下文中，除非另有之处，目标对象的第一位置信息包括目标对象的几何中心的第一位置信息。
这样的实施方式中，在安检过程中获取了待测物的透视图像，并将第一信息与透视图像进行关联。此后，对透视图像中的目标对象进行识别。如所获取的透视图像中显示存在目标对象(违禁物)，则进一步获取待测物的三维外形图像。如所获取的透视图像中不存在目标对象(违禁物)，则完成对待测物的安检。这样，无需对每个待测物均构建三维外形图像，可较为有效地降低数据处理量。
在一个具体实施例中，包括但不限于以待测物中的一点(如集装箱的某一侧壁的中点、货运卡车的前车牌或后车牌中的至少一个)作为原点、获取目标对象相对于原点的三维坐标。
这样的实施方式中，由于原点和待测物的相对位置是固定的，依据该原点建立三维坐标系，有利于标定三维坐标系内的其他坐标点(包括但不限于目标对象和/或移动查看装置)。并且，上述三维外形图像是通过移动查看装置2建立的，移动查看装置2也可作为三维坐标系中的一点，因此，可知移动查看装置2与例如上述的原点之间的相对 位置，进而可获取目标对象相对于移动查看装置2的相对位置(三维坐标)。
根据本公开的一种示意性实施例，获取待测物的透视图像包括但不限于通过X射线扫描待测物获取透视图像。
图3示意性示出了图2中步骤S200的一种实施方法的流程图。
参考图3所示，在步骤S200的一种实施方法中，包括如下子步骤S210至S240。
在步骤S210中，处理器1的识别模块在透视图像中识别待测物中的目标对象，并对目标对象进行标记。这里需要说明的是，识别目标物的具体方法，并不作为本公开的保护要点，任何本领域能够获悉的图像标识方法均可选择适用，不再进行具体展开。
在步骤S220中，处理器1判断目标对象是否具有两个及以上视角的标记。
如果得到两个及以上视角下的标记目标，则在步骤S230中，根据视角和标记目标，确定目标对象在待测物所限定的空间的三维模型内的位置信息。
如果只得到一个视角下的标记目标，则在步骤S240中，通过转换视角，补充标记添加新的标记目标，得到至少一个另一个视角下的标记目标，并返回子步骤S220。
这样的实施方式中，如透视图像是基于一个视角下获取的，则只限定了空间中的一个平面，因此，需从另一个视角下获取其他透视图像，再对该视角下获取的透视图像中的目标对象进行手动标记，由此标定出目标对象在三维模型中的位置信息。如透视图像是基于两个不同的视角下获取的(如相互正交的两个视角)，则可在空间中形成一个三维模型。
根据本公开的另一种示意性实施例，获取待测物的透视图像包括但不限于通过X射线计算机断层扫描待测物获取透视图像。
图4示意性示出了图2中步骤S200的另一种实施方法的流程图。
采用的X射线计算机断层扫描(即CT扫描)，与前述实施方法的区别在于本实施方法中采用CT扫描无需再从多个视角图像获取多个透视图像及得到多个视角下的标记目标。参考图4所示，在步骤S200的另一种实施方法中，具体包括如下子步骤S250至S260。
在步骤S250中，依据待测物的多个断层扫描图像(透视图像)重构三维断层图像，并在三维断层图像中进行标记得到标记目标。
在步骤S260中，依据三维断层图像所建立的第一三维模型，确定目标对象在三维断层图像中的第一位置信息。
这样的实施方式中，X射线计算机断层扫描本身就具备三维成像功能，因此，无需 判定视角的数量，即可获取目标对象的第一位置信息。
图5示意性示出了图2中步骤S300的一种实施方法的流程图。
在步骤S300的一种实施方法中，包括如下子步骤S310至S330。
在步骤S310中，对在步骤S100中利用采集装置3采集过第一信息的待测物识别第二信息。第二信息可以是与第一信息同类的信息，例如集装箱的标号。如下文所述，待测物的第二信息可利用验证模块进行采集。
在步骤S320中，可通过移动查看装置2获取所述待测物的多个外表面的特征。
在步骤S330中，根据所述多个外表面的特征构建所述待测物的三维外形图像。
在步骤S340中，确定移动查看装置相对于三维外形图像的第二位置信息。
根据本公开的一种示意性实施例，步骤S320及S330包括利用移动查看装置2的激光雷达模块采集待测物的外表面的原始点云数据，再对原始点云数据进行配准拼接、去噪简化等处理以获得待测物的实体点云数据，再通过实体点云数据进行建模(包括但不限于几何体创建、平面创建及贴图中的至少一种)，以生成待测物的三维外形图像。
根据本公开的一种示意性实施例，步骤S340包括在待测物上选定一点作为原点，基于该原点及待测物的三维外形图像建立第二三维模型，并将移动查看装置2作为第二三维模型中的一点体现于依据第二三维模型建立的三维坐标系中。
根据本公开的另一种示意性实施例，构建待测物的三维外形图像包括利用三维外形图像获取装置5获取待测物的相对的两侧以及顶部的多个外表面的特征。根据相邻两个外形图像中的共同特征，将多个外形图像拼接成所述待测物的三维外形图像。
在一个具体实施例中，将拍摄装置52，例如摄像机，安装在门架51上，当车辆通过门架51时，利用三个拍摄装置52至少拍摄车辆的相对两侧及顶部的外形图像。车辆的种类可以包括但不限于大型卡车、轻型卡车、运动型多用途汽车(SUV)和轿车等。
这样的实施方式中，三维外形图像获取装置5独立于移动查看装置2以外，可设置于对待测物进行安检的并排位置，以在待测物进行安检的同时一并获取三维外形图像。这样，无需使用者主动构建三维外形图像，可较为有效地降低使用者的工作量。
图6示意性示出了根据本公开实施例的三维外形图像的局部显示效果示意图。
根据本公开的实施例，在车辆经过门架51的情况下，多个拍摄装置52对车辆的外观进行拍摄，并将拍摄的外形图像发送至车辆外观全景拼接装置53中，车辆外观全景拼接装置53根据同一拍摄装置52拍摄的相邻的两个外形图像中的共同特征对多个外形图像进行拼接，从而将多个外形图像拼接成车辆的全景图像，其中，全景图像可以包括 车辆的前表面、后表面、左表面(图6的(a)所示)、右表面(图6的(b)所示)、上表面(图6的(c)所示)和下表面。
根据本公开的一种示意性实施例，由于待测物所限定的容纳空间(如货运卡车的集装箱)相对于待测物(货运卡车)的相对位置是唯一的，因此，依据容纳空间所建立的三维模型相对于待测物的位置是唯一的。将三维模型和待测物的三维外形图像进行关联，依据待测物上的一点(如货运卡车的前车牌的中点)为原点构建三维坐标系。将目标对象在三维模型中的位置信息(如坐标点)映射在三维坐标系中，得到目标对象在三维坐标系中的三维坐标。
这样的实施方式中，透视图像、三维外形图像、待测物及目标对象的位置信息仅存在数据上的关联。这样，所需处理的数据量少，操作较为简便。
根据本公开的另一种示意性实施例，由于可通过透视图像获取三维模型，获取待测物的三维外形图像，因此，可将透视图像及三维外形图像进行拼接以获得拼合图像。获得拼合图像的步骤包括：基于透视图像获取待测物的多个第一识别特征；基于三维外形图像获取待测物的多个第二识别特征；根据第一识别特征和第二识别特征，获取拼合图像；以待测物上的一点作为原点构建拼合图像的三维坐标系；在三维坐标系内获取目标对象的三维坐标。
根据本公开的一种示意性实施例，第一识别特征包括透视图像边缘部分的特征，第二识别特征包括三维外形图像边缘部分的特征。
在本公开的一个具体实施例中，基于透视图像获取车辆的边缘轮廓上的5个特征点作为5个第一识别特征，再基于三维图像获取车辆的边缘轮廓特征上的5个特征点作为5个第二识别特征，将5个第一识别特征和5个第二识别特征进行对准，得到多个拼合图像，选择其中第一识别特征和第二识别特征的对准精度最高的一个拼合图像。
这样的实施方式中，透视图像及三维外形图像被拼接成拼合图像。拼合图像在三维空间中进行表达，目标对象及移动查看装置2均以可视化的形式体现在三维坐标系中(例如，移动查看装置2中以使用者的视角所处的位置作为第二坐标系中的一个坐标点，移动查看装置2中显示目标对象与移动查看装置在三维坐标系中的相对位置)。这样，可具象地体现移动查看装置2与目标对象之间的距离感，以使得使用者的追踪过程更为直观。
根据本公开的一种示意性实施例，将三维坐标关联至移动查看装置2，通过移动查看装置2实时检测移动查看装置2至目标对象的相对位置，以引导移动查看装置2的使 用者靠近目标对象包括：将三维坐标关联至移动查看装置2中；将移动查看装置2的视角设为初始追踪位置，将初始追踪位置至目标对象的方向设为第一追踪方向；移动查看装置2沿第一追踪方向移动，使得移动查看装置2移动后的追踪位置与目标对象之间的距离减小，在减小至预设值时停止移动。在此，预设值可设置为待测物的某一方向上的最小距离的一半，例如集装箱的宽度方向上的长度的一半。通常，在移动查看装置与目标对象之间的距离为预设值时，移动查看装置已较为精准地定位到目标对象，无法或者无需进一步移动。
在一个具体实施例中，将三维坐标关联至移动查看装置2包括将三维坐标系以视频和/或图像形式显示于移动查看装置2中。移动查看装置2进一步被配置成可对显示的三维坐标系进行操作，操作表征为可对三维坐标系的局部或全部区域进行放大、缩小或旋转(包括以三维坐标系中的某一个坐标点在三维空间内旋转和/或绕三维坐标系的某一个轴旋转)。
根据本公开的一种示意性实施例，将三维坐标关联至移动查看装置2，通过移动查看装置2实时检测移动查看装置2至目标对象的相对位置，以引导移动查看装置2的使用者靠近目标对象还包括：在移动查看装置2沿与第一追踪方向不同的第二追踪方向移动，且移动查看装置2移动后的追踪位置与目标对象之间的距离增大时，停止移动，调整移动查看装置2的移动方向至第一追踪方向，沿第一追踪方向移动，使得移动查看装置2移动后的追踪位置与目标对象的距离减小。在此，在移动查看装置2移动后的追踪位置与目标对象之间的距离增大时，移动查看装置2可以振动、响起铃声或出现文字警告等，以提示使用者调整移动查看装置2的移动方向。
这样的实施方式中，在根据本公开实施例的对待测物中的目标对象进行追踪的方法中，移动查看装置2可以显示包括目标对象的三维坐标，使用者可以根据该三维坐标确定移动查看装置2的初始追踪位置至目标对象的追踪方向，之后，使用者可以手持移动查看装置2朝向目标对象的目标对象移动；在移动查看装置2沿第一追踪方向移动时，移动查看装置2移动后的追踪位置与目标对象之间的距离减小，在减小至预设值时停止移动。此后，可采用通常的人工查验确认目标对象的具体位置并确认其是否为违禁物。这样，使用者可以手持移动查看装置2移动并在移动查看装置2上实时显示目标对象的相对三维坐标，可基本上确定目标对象在待测物中的具体位置，从而实现对目标对象的准确追踪和定位，大大缩短了人工追踪的时间，提高了图像的可读性，更加方便快速精准地找到嫌疑物。
根据本公开的另一种总体上的发明构思，还提供一种对待测物中的目标对象进行追踪的系统。
图7示意性示出了根据本公开的一种示意性实施例的用于对待测物中的目标对象进行追踪的系统的方框图。
参考图7所示，对待测物中的目标对象进行追踪的系统包括：处理器1及移动查看装置2。处理器1包括识别模块，被配置成在透视图像中识别并标记待测物中的目标对象，以确定目标对象的第一位置信息。移动查看装置2被配置成获取待测物的三维外形图像、确定移动查看装置2相对于三维外形图像的第二位置信息、根据第一位置信息和第二位置信息获取目标对象相对于移动查看装置2的三维坐标，及将三维坐标关联至移动查看装置2，通过移动查看装置2实时检测移动查看装置2至目标对象的相对位置，以引导移动查看装置2的使用者靠近目标对象。
继续参考图7所示，移动查看装置2包括激光雷达模块。激光雷达模块被配置成采集待测物的外表面的原始点云数据，再对原始点云数据进行配准拼接、去噪简化等处理以获得待测物的实体点云数据，再通过实体点云数据进行建模(包括但不限于几何体创建、平面创建及贴图中的至少一种)，以生成待测物的三维外形图像。更具体地，激光雷达模块包括但不限于激光雷达扫描仪。
在本公开一些实施例中，已经获取的透视图像包括集装箱(包括ISO航空集装箱)、货物卡车之类的待测物的透视图像。这里关于获取的透视图像可以为通过X射线扫描获得的二维透视图像或通过CT扫描获得的三维透视图像。基于一个视角下X射线扫描获得的单个二维透视图像难以确定目标对象，进而需要基于多个视角进行X射线扫描分别获取每个视角下对应的二维透视图像，进而基于多个二维透视图像确定目标对象。这里的多个一般指至少两个。
这样的实施方式中，在根据本公开实施例的基于对待测物中的目标对象进行追踪的系统中，移动查看装置2可以显示目标对象与移动查看装置2的相对位置。这样，使用者可以手持移动查看装置2实现在三维图像上对目标对象的追踪，提高了图像的可读性，方便更加快速精准地找到嫌疑物。
本公开的一种示意性实施例中，用于对待测物中的目标对象进行追踪的系统可选择性地包括扫描装置4，配置为获取例如集装箱(包括ISO航空集装箱)、货物卡车之类的待测物的透视图像。或者也可以是由获取待测物的透视图像后，将待测物的透视图像存储在数据库中，在本次使用时直接调取该存储在数据库中的待测物的透视图像。
图8示意性示出了图7所示的一种示意性实施例的系统的方框图。
参考图8所示，用对待测物中的目标对象进行追踪的系统还包括适用于扫描待测物，以获取待测物的透视图像的扫描装置4。
在本公开的一种示意性实施例中，扫描装置4包括X射线扫描成像设备，适用于获取待测物的至少两个视角的透视图像。
在一些实施例中，移动查看装置2可以包括手持终端设备，例如，包括但不限于，智能手机、平板电脑、膝上型便携计算机和台式计算机等，或者可以为无人驾驶的各种电子设备，包括但不限于无人机、机器人等。
在一个具体实施例中，移动查看装置2选用智能手机，使用者可以在智能手机的手机屏幕查看当前的使用者所在的追踪位置，并将该位置作为初始追踪位置。使用者可以通过上下左右滑动手机屏幕并手动点选想要选择的位置以确定追踪位置，使用者还可以通过两个手指在屏幕上相对、相背滑动实现局部缩放，更加便于查看细节。更具体地，移动查看装置2可以为带有激光雷达扫描仪的智能终端，例如带有LiDAR的iPhone或iPad等。
继续参考图8所示，用于对待测物中的目标对象进行追踪的系统还包括适用于采集待测物的第一信息的采集装置3。更具体地，采集装置3可以为摄像头，例如可见光摄像头。
在一个具体实施例中，摄像头识别到车辆的车头车牌号和/或车尾车牌号(第一信息)。
在另一个具体实施例中，摄像头识别到ISO航空集装箱上的集装箱编号(第一信息)。
在又一个具体实施例中，摄像头识别到车载ISO航空集装箱上的集装箱编号(第一信息)。
继续参考图8所示，用于对待测物中的目标对象进行追踪的系统还包括验证模块。
在本公开的一种示意性实施例中，验证模块和识别模块通讯连接，被配置成验证待测物的第一信息。更具体地，验证模块可以为集成于移动查看装置2中的摄像头及与摄像头通讯连接的图像识别组件(含图像识别软件)，被配置成采集待测物的至少一部分(含第二信息的部分，如车牌)的图像信息，并对采集的图像信息进行识别以识别第二信息，并将其与存储于数据库中的第一信息进行匹配验证。
这样的实施方式中，采集装置3适用于采集待测物的第一信息，验证模块适用于采集待测物的第二信息，并将第二信息和第一信息进行比较，以对第一信息进行验证。
继续参考图8所示，用于对待测物中的目标对象进行追踪的系统还包括处理器1中的调取模块(图中未示出)。调取模块配置为根据待测物的第一信息从数据库中调取与第一信息对应的待测物中的第一位置信息和/或包含有第一位置信息的透视图像。待测物包括集装箱或车辆，待测物的第一信息包括集装箱编号或车辆车牌号。更具体地，调取模块可以为通过管道通信方式由数据库或存储有待测物的相关信息(包括三维模型及三维模型内的位置信息)的云平台调取数据的功能模块。
这样的实施方式中，通过验证模块对第一信息验证后，调取存储于数据库中的与第一信息对应的第一位置信息和/或包含有第一位置信息的透视图像。
图10示意性示出了根据本公开的一种示意性实施例中扫描装置的结构示意图。
参考图10所示，扫描装置4包括第一辐射源40，在一个视角下向待测物，例如集装箱，发射X射线，安装在龙门架54上的两组探测器阵列(图中未示出)接收穿过待测物的X射线，从而得到待测物在一个视角下的透视图像(如图10所述的高度方向)。为了通过第一辐射源40得到待测物在另一个视角下的透视图像，可将待测物转换方向(如将待测物设置成沿长度方向再次通过第一辐射源40的辐射位置，图中未示出)，或者调整第一辐射源40发射射线的角度，或者再设置一组第一辐射源40，以获得另一个视角下的透视图像。
在一些实施例中，识别模块被配置成：在透视图像中识别待测物中的目标对象，并对目标对象进行标记，得到两个及以上视角下的标记目标；以及根据视角和标记目标，确定目标对象在透视图像中的位置信息。
在一些实施例中，识别模块被配置成：在透视图像中识别待测物中的目标对象，并对目标对象进行标记，得到一个视角下的标记目标；通过增加视角，补充标记添加新的标记目标，得到两个以上视角下的标记目标；以及根据视角和标记目标，确定目标对象在待测物中的三维模型的第一位置信息。
在一些实施例中，识别模块被配置成：在透视图像中识别待测物中的目标对象，并对目标对象进行标记得到标记目标；以及根据视角和标记目标，确定目标对象在透视图像中的第一位置信息。
在本公开的另一种示意性实施例中，扫描装置4包括X射线断层扫描设备，适用于获取待测物的三维的透视图像。
在一些实施例中，扫描装置4为基于CT扫描的X射线成像设备。CT扫描是一种利用计算机技术对被测物体断层扫描图像进行重建获得三维断层图像的扫描方式。该扫 描方式通过单一轴面的射线穿透被测物体，根据被测物体各部分对射线的吸收与透过率不同，由计算机根据探测器接收的透过射线进行三维重构成像，由此基于CT扫描的透视图像获取装置仅利用一套射线源和探测器就可以获取三维的透视图像。
图11示意性示出了根据本公开的另一种示意性实施例中扫描装置的结构示意图。
参考图11所示，扫描装置4包括适用于发射X射线的第二辐射源41、校正器42、前准直器43、圆环形旋转架44、安装在圆环形旋转架44上的探测器阵列45和后准直器46、适用于输送集装箱车辆的传送装置47、适用于驱动圆环形旋转架44旋转的驱动装置48、以及适用于制动传送装置47的制动装置49等。
在一个具体实施例中，圆环形旋转架44为一个可以旋转的大型圆环，第二辐射源41和探测器阵列45都固定在该圆环形旋转架44上，圆环形旋转架44旋转进行扫描，得到集装箱的三维透视图像。
在本公开的一种示意性实施例中，识别模块进一步被配置成在透视图像中识别待测物中的目标对象，并对目标对象进行标记，得到至少两个视角下的标记目标。根据视角和标记目标，确定目标对象在待测物所限定的空间的三维模型内的位置信息。
在本公开的一种示意性实施例中，识别模块进一步被配置成在透视图像中识别待测物中的目标对象，并对目标对象进行标记。得到一个视角下的标记目标，通过转换视角，补充标记添加新的标记目标，得到至少两个视角下的标记目标。根据视角和标记目标，确定目标对象在待测物所限定的空间的三维模型内的位置信息。
在一些实施例中，验证模块包括但不限于集成于移动查看装置2中的摄像头及与摄像头通讯连接的图像识别组件(含图像识别软件)。
图9示意性示出了图7所示的另一种示意性实施例的系统的方框图。
参考图9所示，在本公开的一种示意性实施例中，基于对待测物中的目标对象追踪的系统还包括三维外形图像获取装置5。三维外形图像获取装置5配置为获取待测物的三维外形图像，以构建三维坐标系。
继续参考图9所示，用于对待测物中的目标对象进行追踪的系统还包括拼合模块，被配置成基于透视图像和三维外形图像构造拼合图像，并在拼合图像中构建拼合图像的三维坐标系。
在本公开的一些实施例中，拼合模块进一步被配置成基于透视图像获取待测物的多个第一识别特征基于三维外形图像获取待测物的多个第二识别特征。根据第一识别特征和第二识别特征，获取拼合图像。
在本公开的一些实施例中，第一识别特征包括透视图像边缘部分的特征，第二识别特征包括三维外形图像边缘部分的特征。
图12示意性示出了一种示意性实施例的三维外形图像获取装置的结构示意图。
参照图12所示，在本公开的一种示意性实施例中，三维外形图像获取装置包括：门架51、至少三个拍摄装置52和车辆外观全景拼接装置53。门架51包括至少两个侧柱511和/或至少一个横梁512，横梁512的两端分别与侧柱511连接。至少三个拍摄装置52分别设置于相对的两个侧柱511上以及横梁512上，集装箱车辆驶过门架51，拍摄装置52按照预设帧率拍摄车辆的多个外形图像。车辆外观全景拼接装置53与拍摄装置52通信连接，用于根据每个拍摄装置52拍摄的相邻两个外形图像中的共同特征，将多个外形图像拼接成子图像，进而根据多个子图像拼接成待测车辆的三维图像。
作为本公开的一实施例，如果移动查看装置2并未沿第一追踪方向移动，而是沿第二追踪方向移动(第二追踪方向可以是除第一追踪方向外的任意方向，例如与第一追踪方向为相反方向)，移动查看装置2则可以提示使用者改变移动方向。具体地，如果移动查看装置2沿与第一追踪方向不同的第二追踪方向移动，移动查看装置2移动后的追踪位置与目标对象之间的距离增大，则停止移动；进一步地，调整移动查看装置2的移动方向至修改后的第一追踪方向，沿修改后的第一追踪方向移动，使得移动查看装置2移动后的追踪位置与目标对象之间的距离减小，在减小至预设值时停止移动。
在一些实施例中，三维坐标系中的目标对象的位置是固定的，移动查看装置的位置是依据使用者的位置变化而变化的，基于移动查看装置的位置变化可针对使用者的坐标实时更新，并基于更新后的第二坐标点重新获取相对于第一坐标点(目标对象)的三维坐标，以对使用者与目标对象的位置进行修正。
本领域技术人员可以理解，本公开的各个实施例和/或权利要求中记载的特征可以进行多种组合或/或结合，即使这样的组合或结合没有明确记载于本公开中。特别地，在不脱离本公开精神和教导的情况下，本公开的各个实施例和/或权利要求中记载的特征可以进行多种组合和/或结合。所有这些组合和/或结合均落入本公开的范围。
以上对本公开的实施例进行了描述。但是，这些实施例仅仅是为了说明的目的，而并非为了限制本公开的范围。尽管在以上分别描述了各实施例，但是这并不意味着各个实施例中的措施不能有利地结合使用。本公开的范围由所附权利要求及其等同物限定。不脱离本公开的范围，本领域技术人员可以做出多种替代和修改，这些替代和修改都应落在本公开的范围之内。

Claims (23)

  1. 一种对待测物中的目标对象进行追踪的方法，包括：
    对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息；
    利用移动查看装置构建所述待测物的三维外形图像，确定所述移动查看装置相对于所述三维外形图像的第二位置信息；
    根据第一位置信息和第二位置信息，获取所述目标对象相对于移动查看装置的三维坐标；以及
    将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象。
  2. 根据权利要求1所述的方法，还包括获取待测物的第一信息及透视图像。
  3. 根据权利要求2所述的方法，其中，所述利用移动查看装置构建所述待测物的三维外形图像包括：
    识别所述第一信息；以及
    构建与所述第一信息对应的所述待测物的三维外形图像。
  4. 根据权利要求2或3所述的方法，其中，所述获取待测物的透视图像包括通过X射线扫描待测物、获取所述透视图像。
  5. 根据权利要求2或3所述的方法，其中，获取待测物的透视图像包括通过X射线计算机断层扫描待测物、获取三维的所述透视图像。
  6. 根据权利要求4所述的方法，其中，所述对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息包括：
    在所述透视图像中识别并标记所述待测物中的目标对象，得到一个视角下的标记目标；
    通过增加视角，补充标记添加新的标记目标，得到至少一个另一个视角下的标记目标；以及
    根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
  7. 根据权利要求4所述的方法，其中，所述对透视图像中的目标对象进行识别，并确定所述目标对象在待测物中的三维模型的第一位置信息包括：
    在所述透视图像中识别并标记所述待测物中的目标对象，得到至少两个视角下的标记目标；以及
    根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
  8. 根据权利要求1至3中任一所述的方法，其中，所述构建所述待测物的三维外形图像包括：
    利用移动查看装置获取所述待测物的多个外表面的特征；以及
    根据多个所述外表面的特征构建所述待测物的三维外形图像。
  9. 根据权利要求1至3中任一所述的方法，其中，所述根据第一位置信息和第二位置信息，获取所述目标对象相对于所述移动查看装置的三维坐标包括：
    以所述待测物上的一点作为原点构建所述三维外形图像的三维坐标系；
    在三维坐标系中生成所述第一位置信息的第一坐标点；
    在三维坐标系中生成所述第二位置信息的第二坐标点；以及
    获取所述第一坐标点相对于第二坐标点在所述三维坐标系中的三维坐标。
  10. 根据权利要求1至3中任一所述的方法，其中，所述将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象包括：
    将所述三维坐标关联至移动查看装置中；
    将所述移动查看装置的视角设为初始追踪位置，将所述初始追踪位置至所述目标对象的方向设为第一追踪方向；以及
    所述移动查看装置沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象之间的距离减小，在减小至预设值时停止移动。
  11. 根据权利要求10所述的方法，其中，所述将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象还包括：在所述移动查看装置沿与第一追踪方向不 同的第二追踪方向移动，且所述移动查看装置移动后的追踪位置与所述目标对象之间的距离增大时，调整所述移动查看装置的移动方向至第一追踪方向，沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象的距离减小。
  12. 一种用于对待测物中的目标对象进行追踪的系统，包括：
    处理器，包括：
    识别模块，被配置成在透视图像中识别并标记所述待测物中的目标对象，以确定所述目标对象在待测物中的三维模型的第一位置信息；以及
    移动查看装置，被配置成获取所述待测物的三维外形图像；
    其中，所述移动查看装置还被配置成确定所述移动查看装置相对于所述三维外形图像的第二位置信息；根据第一位置信息和第二位置信息，获取所述目标对象相对于移动查看装置的三维坐标；将所述三维坐标关联至移动查看装置，通过所述移动查看装置实时查看移动查看装置至目标对象的相对位置，以引导所述移动查看装置的使用者靠近所述目标对象。
  13. 根据权利要求12所述的系统，还包括扫描装置，适用于扫描所述待测物，以获取所述待测物的透视图像。
  14. 根据权利要求13所述的系统，其中，所述扫描装置包括X射线扫描成像设备，适用于获取所述待测物的至少一个视角的透视图像。
  15. 根据权利要求13所述的系统，其中，所述扫描装置包括X射线断层扫描设备，适用于获取所述待测物的三维的透视图像。
  16. 根据权利要求12至15中任一所述的系统，还包括采集装置，适用于采集所述待测物的第一信息。
  17. 根据权利要求16所述的系统，其中，所述移动查看装置还包括验证模块，被配置成验证所述待测物的第一信息。
  18. 根据权利要求17所述的系统，其中，所述处理器还包括调取模块，配置为从数据库中调取与第一信息对应的待测物中的目标对象的第一位置信息；
    所述待测物包括集装箱或车辆；所述待测物的第一信息包括集装箱编号或车辆车牌号。
  19. 根据权利要求12至14中任一所述的系统，其中，所述识别模块进一步被配置成：
    在所述透视图像中识别所述待测物中的目标对象，并对所述目标对象进行标记，得到一个视角下的标记目标；
    通过增加视角，补充标记添加新的标记目标，得到至少两个视角下的标记目标；以及
    根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
  20. 根据权利要求12至14中任一所述的系统，其中，所述识别模块进一步被配置成：
    在所述透视图像中识别所述待测物中的目标对象，并对所述目标对象进行标记，得到至少两个视角下的标记目标；以及
    根据所述视角和所述标记目标，确定所述目标对象在所述待测物所限定的空间的三维模型内的第一位置信息。
  21. 根据权利要求12至15中任一项所述的系统，其中，所述移动查看装置包括激光雷达模块，被配置成获取所述待测物的多个外表面的特征，并根据所述多个外表面的所述特征构建所述待测物的三维外形图像。
  22. 根据权利要求12至15中任一项所述的系统，其中，所述移动查看装置进一步被配置成：
    以所述待测物上的一点作为原点构建所述三维外形图像的三维坐标系；
    在三维坐标系中生成所述第一位置信息的第一坐标点；
    在三维坐标系中生成所述第二位置信息的第二坐标点；以及
    获取所述第一坐标点相对于第二坐标点在所述三维坐标系中的三维坐标。
  23. 根据权利要求12至15中任一所述的系统，其中，所述移动查看装置被配置成：
    在沿与第一追踪方向不同的第二追踪方向移动，且所述移动查看装置移动后的追踪位置与所述目标对象对应之间的距离增大时，停止移动；以及
    调整所述移动查看装置的移动方向至第一追踪方向，沿所述第一追踪方向移动，使得所述移动查看装置移动后的追踪位置与所述目标对象之间的距离减小，在减小至预设值时停止移动。
PCT/CN2023/104626 2022-07-01 2023-06-30 对待测物中的目标对象进行追踪的方法及系统 WO2024002340A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210774556.0A CN116224355A (zh) 2022-07-01 2022-07-01 对待测物中的目标对象进行追踪的方法及系统
CN202210774556.0 2022-07-01

Publications (1)

Publication Number Publication Date
WO2024002340A1 true WO2024002340A1 (zh) 2024-01-04

Family

ID=86579164

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/104626 WO2024002340A1 (zh) 2022-07-01 2023-06-30 对待测物中的目标对象进行追踪的方法及系统

Country Status (2)

Country Link
CN (1) CN116224355A (zh)
WO (1) WO2024002340A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116224355A (zh) * 2022-07-01 2023-06-06 同方威视技术股份有限公司 对待测物中的目标对象进行追踪的方法及系统
CN117590479A (zh) * 2022-08-08 2024-02-23 同方威视技术股份有限公司 嫌疑物品定位系统和定位方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101375805A (zh) * 2007-12-29 2009-03-04 清华大学深圳研究生院 一种计算机辅助引导电子内窥镜操作的方法和系统
US20130253824A1 (en) * 2009-05-11 2013-09-26 Acer Incorporated Object guiding method, mobile viewing system and augmented reality system
CN105992153A (zh) * 2015-06-03 2016-10-05 乐卡汽车智能科技（北京）有限公司 一种定位信息展示方法及装置
CN109187606A (zh) * 2018-08-13 2019-01-11 同济大学 一种复合式违禁品智能化精密检测装备和检测方法
CN111142067A (zh) * 2019-08-27 2020-05-12 广东小天才科技有限公司 一种移动设备的立体定位方法及移动设备
CN112069841A (zh) * 2020-07-24 2020-12-11 华南理工大学 新型x光违禁品包裹跟踪方法及装置
CN116224355A (zh) * 2022-07-01 2023-06-06 同方威视技术股份有限公司 对待测物中的目标对象进行追踪的方法及系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101375805A (zh) * 2007-12-29 2009-03-04 清华大学深圳研究生院 一种计算机辅助引导电子内窥镜操作的方法和系统
US20130253824A1 (en) * 2009-05-11 2013-09-26 Acer Incorporated Object guiding method, mobile viewing system and augmented reality system
CN105992153A (zh) * 2015-06-03 2016-10-05 乐卡汽车智能科技（北京）有限公司 一种定位信息展示方法及装置
CN109187606A (zh) * 2018-08-13 2019-01-11 同济大学 一种复合式违禁品智能化精密检测装备和检测方法
CN111142067A (zh) * 2019-08-27 2020-05-12 广东小天才科技有限公司 一种移动设备的立体定位方法及移动设备
CN112069841A (zh) * 2020-07-24 2020-12-11 华南理工大学 新型x光违禁品包裹跟踪方法及装置
CN116224355A (zh) * 2022-07-01 2023-06-06 同方威视技术股份有限公司 对待测物中的目标对象进行追踪的方法及系统

Also Published As

Publication number Publication date
CN116224355A (zh) 2023-06-06

Similar Documents

Publication Publication Date Title
WO2024002340A1 (zh) 对待测物中的目标对象进行追踪的方法及系统
EP3208774B1 (en) Methods for localization using geotagged photographs and three-dimensional visualization
JP5480914B2 (ja) 点群データ処理装置、点群データ処理方法、および点群データ処理プログラム
JP5467404B2 (ja) ３ｄ撮像システム
US8639025B2 (en) Measurement apparatus and control method
US9470511B2 (en) Point-to-point measurements using a handheld device
JP5580164B2 (ja) 光学情報処理装置、光学情報処理方法、光学情報処理システム、光学情報処理プログラム
US6381302B1 (en) Computer assisted 2D adjustment of stereo X-ray images
US8224121B2 (en) System and method for assembling substantially distortion-free images
JP2018028535A (ja) 検査ステーションに配置された車両の最外寸法を測定するための方法及びシステム
CN104567758B (zh) 立体成像系统及其方法
CN106659448A (zh) 用于配置x射线成像系统的方法和系统
CN105849772B (zh) 检查系统及方法
JP2012037491A (ja) 点群位置データ処理装置、点群位置データ処理システム、点群位置データ処理方法、および点群位置データ処理プログラム
JP2010523950A (ja) 透過する放射線を用いて計測対象の三次元画像を生成するための方法および計測装置
CN112254902B (zh) 基于激光和可见光扫描生成立体激光点云图的方法及装置
JP2010540893A (ja) Ｘ線容積撮影による画像再構成方法
US20190129045A1 (en) Radiation imaging apparatus
JP2021056017A (ja) 合成処理装置、合成処理システム及び合成処理方法
WO2021226716A1 (en) System and method for discrete point coordinate and orientation detection in 3d point clouds
KR20040010297A (ko) Ｃｔ 장치, ｃｔ 촬상 방법, 및 ｃｔ 촬상 서비스 방법
EP3112852B1 (en) Method for positioning target in three-dimensional ct image and security check system
KR102285337B1 (ko) 엑스선 장치의 캘리브레이션 방법 및 장치
GB2551970A (en) Determining the position of an object in a scene
Zhu et al. Gamma/X-ray linear pushbroom stereo for 3D cargo inspection

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: 23830503

Country of ref document: EP

Kind code of ref document: A1