CN114511592B - Personnel track tracking method and system based on RGBD camera and BIM system - Google Patents
Personnel track tracking method and system based on RGBD camera and BIM system Download PDFInfo
- Publication number
- CN114511592B CN114511592B CN202210071107.XA CN202210071107A CN114511592B CN 114511592 B CN114511592 B CN 114511592B CN 202210071107 A CN202210071107 A CN 202210071107A CN 114511592 B CN114511592 B CN 114511592B
- Authority
- CN
- China
- Prior art keywords
- rgbd
- person
- target
- frame
- personnel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000013507 mapping Methods 0.000 claims abstract description 6
- 238000001514 detection method Methods 0.000 claims description 93
- 230000003287 optical effect Effects 0.000 claims description 36
- 238000010606 normalization Methods 0.000 claims description 17
- 238000004422 calculation algorithm Methods 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000011426 transformation method Methods 0.000 abstract 1
- 238000010276 construction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Abstract
The invention provides a personnel track tracking method and system based on an RGBD camera and a BIM system, wherein the method is based on detecting and tracking a target personnel in an RGBD image, calculating the actual position (x, y) of a central point of the target personnel in a target area which can be shot by the RGBD camera through a coordinate transformation method, calculating the position of the target personnel in an actual scene by combining an RGBD camera number and a personnel ID number, drawing the motion track of the target personnel in the actual scene according to the position of the target personnel in the actual scene, and mapping the position and the motion track of the target personnel in the actual scene to the BIM system for real-time display. The personnel track tracking method and system based on the RGBD camera and the BIM system effectively solve the technical problems that the positions of the personnel in the actual scene cannot be positioned and the motion tracks of the personnel in the actual scene cannot be drawn and displayed in real time in the prior art.
Description
Technical Field
The invention relates to the field of computer vision research, in particular to a personnel track tracking method and system based on an RGBD camera and a BIM system.
Background
Along with the development of the computer vision field, the application of personnel track analysis is more and more widespread, most of personnel track tracking methods in the prior art track personnel tracks the personnel track based on RGB image information, and although the technical scheme can determine that the target personnel appear in a target area monitored by a camera, the position of the target personnel in an actual scene cannot be positioned and displayed, and the movement track of the target personnel in the actual scene cannot be accurately depicted, so that the requirements of acquiring the position of the target personnel in the actual scene and displaying the position of the target personnel are higher and higher in the fields related to BIM construction, such as intelligent security, intelligent communities, intelligent buildings and the like. In view of this, the present application has been proposed.
Disclosure of Invention
The invention mainly aims to solve the problems and the defects, provides a person track tracking method based on an RGBD camera and a BIM system, which is used for calculating the position of a person in an actual scene based on the RGBD camera and drawing the motion track of the person in a BIM system for real-time display.
In order to achieve the above purpose, the present invention firstly provides a personnel track tracking method based on an RGBD camera and a BIM system, which has the technical scheme that:
A person trajectory tracking method based on an RGBD camera and a BIM system, the method comprising,
S1, acquiring an RGBD image of a target area by adopting an RGBD camera, and distributing an RGBD camera number for the RGBD camera;
s2, detecting the RGBD image, obtaining a personnel detection frame, and distributing a personnel ID number for the personnel detection frame;
s3, tracking a target person in the person detection frame, acquiring the person detection frame of the target person in the RGBD image of each frame, and distributing the same person ID number for the person detection frame of the same target person in the RGBD image of each frame;
And S4, calculating the position of the target person in the actual scene in each frame of RGBD image based on the person detection frame in each frame of RGBD image and the depth information in the RGBD image, and mapping the position in the actual scene and the motion trail of the target person to a BIM system for real-time display.
Further, in the step S4, an actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area is calculated by a coordinate transformation means, and then the position of the target person in each frame of RGBD image in the actual scene is calculated according to the RGBD camera number and the person ID number.
Further, in the step S4, the calculating, by means of coordinate transformation, an actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area is to convert the RGBD image to a top view, taking an upper left corner of the RGBD image as an origin, taking right and downward directions as positive directions of an x axis and a y axis, respectively, and establishing a planar rectangular coordinate system, and calculating an actual position (x, y) of the center point of the target person in the target area,
The calculation formula of x is:
when the center point of the target person is located above the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
When the center point of the target person is located below the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
In the method, in the process of the invention,
D: the person detects the actual straight line distance between the center point of the frame and the optical center of the RGBD camera,
D1: the person detects the vertical pixel distance of the center point of the frame from the optical center of the RGBD camera,
D2: the person detects the horizontal pixel distance of the center point of the frame from the optical center of the RGBD camera,
Θ 1: the angle of the RGBD camera optical axis in the vertical direction,
F: the pixel focal length of the RGBD camera,
Y1: the vertical distance of the upper edge of the RGBD image to the line where the camera optical axis is located.
Further, in the step S2, the method for detecting the RGBD image and obtaining the human detection frame includes:
S21, adjusting the resolution ratio of the RGBD image of a single frame, and carrying out normalization processing on each pixel value of the RGBD image of the single frame;
And S22, performing feature extraction and classification processing on the RGBD image subjected to normalization processing by adopting a pedestrian detection network, and screening out all personnel detection frames.
Further, the formula of the normalization process in the step S21 is as follows:
In the method, in the process of the invention,
U: the pixel values after the normalization are used,
V: the pixel values before the normalization are compared with each other,
V min: the smallest pixel value in the RGBD image of a single frame,
V max: and the maximum pixel value in the RGBD image of a single frame.
Further, the pedestrian detection network includes, but is not limited to, one of an SSD network, RCNN network, and YOLO series network.
Further, in the step S3, a multi-target tracking algorithm is adopted to track the target person in the person detection frame, which includes the following steps:
S31, predicting the position of the personnel detection frame in the i-1 frame in the RGBD image of the i frame to obtain a predicted position, and matching the predicted position with all the personnel detection frames in the RGBD image of the i frame to obtain a target detection frame matched with the predicted position;
S32, intercepting a target image at a position corresponding to the target detection frame from an ith frame RGBD image, and adjusting the size of the target image;
s33, extracting 512-dimensional features of all people in the target image, and calculating the similarity between the 512-dimensional features of each person and the 512-dimensional features of the target person in the i-1 frame RGBD image by adopting a cosine distance;
S34, matching the target person in the i-1 th frame RGBD image with the person in the target image by using a Hungary algorithm, wherein the similarity accords with the definition of a matching condition to be the target person, and acquiring a person detection frame of the target person in the i-1 th frame RGBD image;
Wherein i is not less than 2.
Further, the multi-target tracking algorithm includes, but is not limited to, one of DeepSORT tracking algorithm and pysot tracking algorithm.
The invention further provides a personnel track tracking system based on an RGBD camera and a BIM system, which has the technical scheme that:
A personnel track tracking system based on RGBD camera and BIM system comprises a camera shooting module, a pedestrian detection module, a target tracking module, a positioning module and a BIM system module, wherein,
The camera shooting module is used for acquiring RGBD images of a target area by adopting an RGBD camera and distributing RGBD camera numbers to the RGBD camera;
The pedestrian detection module is used for detecting the RGBD image, acquiring a personnel detection frame and distributing a personnel ID number for the personnel detection frame;
The target tracking module is used for tracking target personnel in the personnel detection frames, acquiring the personnel detection frames of the target personnel in each frame of RGBD image, and distributing the same personnel ID number for the personnel detection frames of the same target personnel;
The positioning module calculates the actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area, and fuses the RGBD camera number, the person ID number and the actual position (x, y) of the center point of the target person in a single frame of RGBD image into one piece of position information to be sent to the BIM system module;
and the BIM system module is used for determining the position of the target person in the actual scene according to the position information, and mapping the position in the actual scene to the front end of the BIM system according to the time sequence for display.
The invention further provides computer equipment, which has the technical scheme that:
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the memory having stored thereon program code for a RGBD camera and BIM system based person trajectory tracking method as described above, the processor implementing at least the steps of the RGBD camera and BIM system based person trajectory tracking method as described above when executing the computer program.
In summary, compared with the prior art, the personnel track tracking method and system based on the RGBD camera and the BIM system provided by the invention calculate the actual position (x, y) of the center point of the target personnel in the target area of the RGBD camera based on the RGB image information and the depth information contained in the RGBD image, then determine the position of the personnel in the actual scene according to the RGBD camera number and the personnel ID number, map the position and the motion track of the target personnel drawn according to the position to the BIM system for real-time display, and effectively solve the technical problems that the position of the personnel in the actual scene cannot be positioned and the motion track of the personnel in the actual scene cannot be drawn and displayed in the prior art; in step S3, the person in the person detection frame is tracked based on the RGB image information contained in the RGBD image, and the depth information is not required to be combined, so that the operation amount is reduced while the tracking accuracy of the person track is ensured, and the purpose of low power consumption is achieved.
Description of the drawings:
Fig. 1: the invention relates to a personnel track tracking method based on an RGBD camera and a BIM system, which comprises the following steps of;
fig. 2: a flow chart of the person detection step of the present invention;
fig. 3: a flow chart of the person tracking step of the present invention;
fig. 4: a flow chart of the personnel position calculation of the present invention;
fig. 5: the invention relates to a schematic diagram of the relative positions of RGDB cameras and personnel under the front view angle;
fig. 6: the invention relates to a schematic diagram of the relative positions of a RGDB camera and a person in a overlook view.
Detailed Description
The invention is described in further detail below with reference to the drawings and the detailed description.
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment provides a personnel track tracking method and system based on an RGBD camera and a BIM system, which can accurately track the track of a personnel, simultaneously position the personnel in an actual scene, and display the position of the personnel in the actual scene at the front end of the BIM system in real time, so as to better meet the requirements of personnel tracking in the fields related to BIM construction such as intelligent security, intelligent communities, intelligent buildings and the like, and as shown in fig. 1, the personnel track tracking method based on the RGBD camera and the BIM system mainly comprises the processes of detecting the personnel in an RGBD image, tracking the personnel, calculating the position of the personnel in the actual scene and displaying the position of the personnel in the actual scene in the BIM system, and specifically comprises the following steps:
S1, acquiring RGBD images of a target area by adopting an RGBD camera, and allocating RGBD camera numbers for the RGBD camera. The RGBD image includes RGB image information for detection and tracking of a person and depth information combined for position calculation of the person in an actual scene.
S2, detecting the RGBD image, acquiring a personnel detection frame, and distributing personnel ID numbers for the personnel detection frame.
In step S2, RGB image information included in the RGBD image is detected in units of frames, and a person detection frame is obtained, see fig. 2, which specifically includes the steps of:
s21, adjusting the resolution ratio of the single-frame RGBD image, and carrying out normalization processing on each pixel value of the single-frame RGBD image;
Specifically, in order to quickly and accurately perform personnel detection, the calculated amount is reduced, the resolution of each frame of RGBD image is uniformly adjusted to 640 x 640, then normalization processing is performed on each pixel value of the RGBD image after resolution adjustment, and the normalized RGBD image is sent to YOLOv pedestrian detection network for personnel detection. Wherein, the normalization formula is:
In the method, in the process of the invention,
U: the pixel values after the normalization are used,
V: the pixel values before the normalization are compared with each other,
V min: the smallest pixel value in a single frame RGBD image,
V max: the largest pixel value in a single frame RGBD image.
And S22, performing feature extraction and classification processing on the RGBD image subjected to normalization processing by adopting YOLOv pedestrian detection network, and screening out all personnel detection frames.
And YOLOv, after receiving the RGB image subjected to normalization processing, the pedestrian detection network extracts and classifies the features of the RGBD image to obtain three feature images of 80 x 255, 40 x 255 and 20 x 255, processes the three feature images to obtain detection frame coordinates and class labels of all targets in the RGBD image, and finally screens out all personnel detection frames according to the class labels.
In the present invention, the pedestrian detection network includes, but is not limited to, one of an SSD network, RCNN network, and YOLO series network, and the embodiment uses YOLOv of the YOLO series network to obtain the person detection frames, but the present invention is not limited in any way, and in practical application, the pedestrian detection network can be used only to screen out all the detection frames with a category label of "person".
S3, tracking the target person in the person detection frame, acquiring the person detection frame of the target person in each frame of RGBD image, and distributing the same person ID number for the person detection frame of the same target person in each frame of RGBD image.
In step S3, based on the person detection frame acquired in step S2 and RGB image information included in the original RGBD image, a DeepSORT multi-target tracking algorithm is used to track a target person in the person detection frame, see fig. 3, including the following steps:
S31, predicting the position of a person detection frame in the i-1 th frame in the i-th frame RGBD image through a Kalman filtering algorithm to obtain a predicted position, and matching the predicted position with all the person detection frames in the i-th frame RGBD image to obtain a target detection frame with highest matching degree;
S32, intercepting a target image at a position corresponding to the target detection frame in the ith frame RGBD image, adjusting the size of the target image to 128 x 64, and sending the adjusted target image to a feature extractor;
S33, extracting 512-dimensional features of all people in the target image in a feature extractor, calculating the similarity between the 512-dimensional features of each person and the 512-dimensional features of the target person in a person detection frame in an i-1 frame RGBD image by adopting a cosine distance, and selecting a small residual error network as a feature extractor to obtain the 512-dimensional features as a preferable choice;
S34, performing feature matching on a target person in the i-1 frame RGBD image and a person in the intercepted target image by using a Hungary algorithm, if the similarity value is more than or equal to 0.4, taking the person in the target image corresponding to the maximum similarity value as a matching target and defining the person as a target person, thereby acquiring a person detection frame of the target person in the i frame RGBD image, and distributing the same person ID number for the person detection frame; if the similarity value is less than 0.4, the matching fails, and the person is tracked by adopting a cross-camera.
Wherein i is not less than 2.
In step S3, the personnel in the personnel detection frame is tracked based on the RGB image information, and the depth information is not required to be combined, so that the operation amount is reduced while the personnel track tracking accuracy is ensured, and the purpose of low power consumption is achieved.
It should be noted that, in the present invention, the multi-target tracking algorithm includes, but is not limited to, one of DeepSORT tracking algorithm and pysot tracking algorithm.
And S4, calculating the position of the center point of the target person in each frame of RGBD image in the actual scene based on the person detection frame of the target person in each frame of RGBD image and the depth information in the RGBD image, and mapping the position in the actual scene and the motion trail of the target person to a BIM system for real-time display.
In step S4, the actual position (x, y) of the center point of the target person in the target area in each frame of RGBD image is calculated through the coordinate transformation means, then the area where the target person appears in the actual scene is determined according to the RGBD camera number, then the position of the center point of the target person in the actual scene is determined according to the actual position (x, y) of the center point of the target person in the target area and the person ID number, namely, the position of the center point of the target person in the actual scene is mapped to the BIM system for real-time display, and the positions of the center point of the target person in the actual scene are connected according to the time sequence, so that the motion trail of the target person can be drawn and displayed in the BIM system.
Referring to fig. 4, the specific process of calculating the actual position (x, y) of the center point of the target person in each frame RGBD image in the target area by the coordinate transformation means is to calculate the value of the included angle θ2 between the line between the center point of the person detection frame and the optical center of the RGBD camera and the optical axis of the RGBD camera, the value of the horizontal distance w between the center point of the person detection frame and the optical center of the RGBD camera, and then calculate the value of the included angle θ between the line between the center point of the target person and the optical center of the RGBD camera and the horizontal direction of the optical axis of the RGBD camera in the top view, and then calculate the actual position (x, y) of the center point of the target person in the target area according to the value of θ and the value of w. Specifically, as shown in fig. 5, firstly, converting an RGBD image into a value of an included angle θ2 between a connecting line of a center point of a human detection frame and an optical center of an RGBD camera and an optical axis of the RGBD camera and a value of a horizontal distance w between the center point of the human detection frame and the optical center of the RGBD camera, wherein a calculation formula of θ2 is as follows: The calculation formula of w is: Then as shown in fig. 6, converting the RGBD image to a top view angle, taking the upper left corner of the RGBD image as an origin, taking the right and downward directions as positive directions of an x axis and a y axis, respectively, establishing a plane rectangular coordinate system, and calculating the actual position (x, y) of a person in a target area, specifically, firstly calculating the value of an included angle θ between the connecting line of the center point of the person and the optical center of the RGBD camera and the horizontal direction of the optical axis of the RGBD camera, wherein the calculation formula of θ is as follows: And then calculating the actual position (x, y) of the center point of the target person in the target area according to the value of the horizontal distance w between the center point of the person detection frame and the optical center of the RGBD camera and the value of the included angle theta between the connecting line of the center point of the target person and the optical center of the RGBD camera and the horizontal direction of the optical axis of the RGBD camera, wherein,
The calculation formula of x is:
when the center point of the target person is located above the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
When the center point of the target person is located below the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
when the center point of the target person is located on the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows: y=y1 and,
In the method, in the process of the invention,
D: the person detects the actual straight line distance between the center point of the frame and the optical center of the RGBD camera,
D1: the person detects the vertical pixel distance of the center point of the frame from the optical center of the RGBD camera,
D2: the person detects the horizontal pixel distance of the center point of the frame from the optical center of the RGBD camera,
Θ 1: the angle of the RGBD camera optical axis in the vertical direction,
F: the pixel focal length of the RGBD camera,
Y1: the vertical distance from the upper edge of the RGBD image to the line where the camera optical axis is located,
Y2: the vertical distance from the center point of the target person to the straight line where the optical axis of the camera is located.
In order to facilitate understanding and explanation of the present invention, the present embodiment describes the calculation step of the actual position (x, y) of the center point of the target person in the target area by converting the RGBD image into the front view angle and the top view angle.
The invention further provides a personnel track tracking system based on an RGBD camera and a BIM system, which comprises a camera shooting module, a pedestrian detection module, a target tracking module, a positioning module and a BIM system module, wherein the camera shooting module acquires an RGBD image of a target area by adopting the RGBD camera, assigns an RGBD camera number for the RGBD camera, and transmits the RGBD image containing RGB image information and depth information to the pedestrian detection module; the pedestrian detection module is used for detecting the RGBD image, acquiring a personnel detection frame, distributing personnel ID numbers for the personnel detection frame, and then conveying all acquired personnel detection frames and original RGBD images to the target tracking module; the target tracking module is used for tracking target personnel in the personnel detection frames, acquiring the personnel detection frames of the target personnel in each frame of RGBD image, and distributing the same personnel ID number for the personnel detection frames of the same target personnel; the positioning module is used for calculating the actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area by combining the RGB image information and the depth information in the RGBD image, and fusing the RGBD camera number, the person ID number and the actual position (x, y) of the center point of the target person in a single frame of RGBD image into one piece of position information to be sent to the BIM system module; and the background of the BIM system module calculates the position of the target person in the actual scene according to the position information, arranges the positions in the actual scene according to the time sequence, draws the motion trail of the target person in real time and displays the motion trail at the front end of the BIM system module.
The invention further provides a computer device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the memory has stored thereon program code for a person trajectory tracking method based on an RGBD camera and a BIM system as described above, and wherein the processor when executing the computer program at least implements the steps of the person trajectory tracking method based on an RGBD camera and a BIM system as described above.
In summary, compared with the prior art, the personnel track tracking method and system based on the RGBD camera and the BIM system provided by the invention calculate the actual position (x, y) of the center point of the target personnel in the target area of the RGBD camera based on the RGB image information and the depth information contained in the RGBD image, then determine the position of the personnel in the actual scene according to the RGBD camera number and the personnel ID number, map the position and the motion track of the target personnel drawn according to the position to the BIM system for real-time display, and effectively solve the technical problems that the position of the personnel in the actual scene cannot be positioned and the motion track of the personnel in the actual scene cannot be drawn and displayed in the prior art; in step S3, the person in the person detection frame is tracked based on the RGB image information contained in the RGBD image, and the depth information is not required to be combined, so that the operation amount is reduced while the tracking accuracy of the person track is ensured, and the purpose of low power consumption is achieved.
As mentioned above, similar technical solutions can be derived in combination with the presented solution content. However, any simple modification, equivalent variation and modification of the above embodiments according to the technical substance of the present invention still fall within the scope of the technical solution of the present invention.
Claims (9)
1. A personnel track tracking method based on an RGBD camera and a BIM system is characterized in that: the method may include the steps of,
S1, acquiring an RGBD image of a target area by adopting an RGBD camera, and distributing an RGBD camera number for the RGBD camera;
s2, detecting the RGBD image, obtaining a personnel detection frame, and distributing a personnel ID number for the personnel detection frame;
s3, tracking a target person in the person detection frame, acquiring the person detection frame of the target person in the RGBD image of each frame, and distributing the same person ID number for the person detection frame of the same target person in the RGBD image of each frame;
S4, calculating the position of the target person in the actual scene in each frame of RGBD image based on the person detection frame in each frame of RGBD image and the depth information in the RGBD image, and mapping the position in the actual scene and the motion trail of the target person into a BIM system for real-time display;
In the step S4, the actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area is calculated by means of coordinate transformation, in which the actual position (x, y) of the center point of the target person in the target area is calculated by taking the upper left corner of the RGBD image as the origin and taking the right and downward directions as the positive directions of the x-axis and the y-axis, respectively, and a planar rectangular coordinate system is established,
The calculation formula of x is:
when the center point of the target person is located above the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
When the center point of the target person is located below the straight line where the optical axis of the RGBD camera is located, the calculation formula of y is as follows:
In the method, in the process of the invention,
D: the person detects the actual straight line distance between the center point of the frame and the optical center of the RGBD camera,
D 1: the person detects the vertical pixel distance of the center point of the frame from the optical center of the RGBD camera,
D 2: the person detects the horizontal pixel distance of the center point of the frame from the optical center of the RGBD camera,
Θ 1: the angle of the RGBD camera optical axis in the vertical direction,
F: the pixel focal length of the RGBD camera,
Y1: the vertical distance of the upper edge of the RGBD image to the line where the camera optical axis is located.
2. The person trajectory tracking method based on an RGBD camera and BIM system as claimed in claim 1, wherein: in the step S4, the actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area is calculated by the coordinate transformation means, and then the position of the target person in each frame of RGBD image in the actual scene is calculated according to the RGBD camera number and the person ID number.
3. The person trajectory tracking method based on an RGBD camera and BIM system as claimed in claim 1, wherein: in the step S2, the method for detecting the RGBD image and obtaining the human detection frame includes:
S21, adjusting the resolution ratio of the RGBD image of a single frame, and carrying out normalization processing on each pixel value of the RGBD image of the single frame;
And S22, performing feature extraction and classification processing on the RGBD image subjected to normalization processing by adopting a pedestrian detection network, and screening out all personnel detection frames.
4. A person tracking method based on RGBD camera and BIM system as claimed in claim 3, wherein: the formula of the normalization process in the step S21 is:
In the method, in the process of the invention,
U: the pixel values after the normalization are used,
V: the pixel values before the normalization are compared with each other,
V min: the smallest pixel value in the RGBD image of a single frame,
V max: and the maximum pixel value in the RGBD image of a single frame.
5. A person tracking method based on RGBD camera and BIM system as claimed in claim 3, wherein: the pedestrian detection network includes, but is not limited to, one of an SSD network, RCNN network, and YOLO series network.
6. The person trajectory tracking method based on an RGBD camera and BIM system as claimed in claim 1, wherein: in the step S3, a multi-target tracking algorithm is adopted to track a target person in the person detection frame, and the method includes the following steps:
S31, predicting the position of the personnel detection frame in the i-1 frame in the RGBD image of the i frame to obtain a predicted position, and matching the predicted position with all the personnel detection frames in the RGBD image of the i frame to obtain a target detection frame matched with the predicted position;
S32, intercepting a target image at a position corresponding to the target detection frame from an ith frame RGBD image, and adjusting the size of the target image;
s33, extracting 512-dimensional features of all people in the target image, and calculating the similarity between the 512-dimensional features of each person and the 512-dimensional features of the target person in the i-1 frame RGBD image by adopting a cosine distance;
S34, matching the target person in the i-1 th frame RGBD image with the person in the target image by using a Hungary algorithm, wherein the similarity accords with the definition of a matching condition to be the target person, and acquiring a person detection frame of the target person in the i-1 th frame RGBD image;
Wherein i is not less than 2.
7. The person trajectory tracking method based on an RGBD camera and BIM system as claimed in claim 6, wherein: the multi-target tracking algorithm includes, but is not limited to, one of DeepSORT tracking algorithm and pysot tracking algorithm.
8. A tracking system employing the RGBD camera and BIM system based person tracking method of any of claims 1 to 7, characterized in that: comprises a camera shooting module, a pedestrian detection module, a target tracking module, a positioning module and a BIM system module, wherein,
The camera shooting module is used for acquiring RGBD images of a target area by adopting an RGBD camera and distributing RGBD camera numbers to the RGBD camera;
The pedestrian detection module is used for detecting the RGBD image, acquiring a personnel detection frame and distributing a personnel ID number for the personnel detection frame;
The target tracking module is used for tracking target personnel in the personnel detection frames, acquiring the personnel detection frames of the target personnel in each frame of RGBD image, and distributing the same personnel ID number for the personnel detection frames of the same target personnel;
The positioning module calculates the actual position (x, y) of the center point of the target person in each frame of RGBD image in the target area, and fuses the RGBD camera number, the person ID number and the actual position (x, y) of the center point of the target person in a single frame of RGBD image into one piece of position information to be sent to the BIM system module;
and the BIM system module is used for determining the position of the target person in the actual scene according to the position information, and mapping the position in the actual scene to the front end of the BIM system according to the time sequence for display.
9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized by: the memory has stored thereon program code for a RGBD camera and BIM system based person trajectory tracking method according to any one of claims 1 to 7, and the processor when executing the computer program implements at least the steps of the RGBD camera and BIM system based person trajectory tracking method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210071107.XA CN114511592B (en) | 2022-01-21 | Personnel track tracking method and system based on RGBD camera and BIM system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210071107.XA CN114511592B (en) | 2022-01-21 | Personnel track tracking method and system based on RGBD camera and BIM system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114511592A CN114511592A (en) | 2022-05-17 |
| CN114511592B true CN114511592B (en) | 2024-07-05 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106570883A (en) * | 2016-10-25 | 2017-04-19 | 长安大学 | People counting method based on RGB-D camera |
| CN106599776A (en) * | 2016-10-25 | 2017-04-26 | 长安大学 | People counting method based on trajectory analysis |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106570883A (en) * | 2016-10-25 | 2017-04-19 | 长安大学 | People counting method based on RGB-D camera |
| CN106599776A (en) * | 2016-10-25 | 2017-04-26 | 长安大学 | People counting method based on trajectory analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021196294A1 (en) | Cross-video person location tracking method and system, and device | |
| CN110674746B (en) | Method and device for realizing high-precision cross-mirror tracking by using video spatial relationship assistance, computer equipment and storage medium | |
| US11393103B2 (en) | Target tracking method, device, system and non-transitory computer readable medium | |
| Sidla et al. | Pedestrian detection and tracking for counting applications in crowded situations | |
| CN109887040B (en) | Moving target active sensing method and system for video monitoring | |
| WO2022135511A1 (en) | Method and apparatus for positioning moving object, and electronic device and storage medium | |
| CN110400352B (en) | Camera calibration with feature recognition | |
| US20150138310A1 (en) | Automatic scene parsing | |
| US20220180534A1 (en) | Pedestrian tracking method, computing device, pedestrian tracking system and storage medium | |
| CN115439424A (en) | Intelligent detection method for aerial video image of unmanned aerial vehicle | |
| CN109359577B (en) | System for detecting number of people under complex background based on machine learning | |
| CN112085534B (en) | Attention analysis method, system and storage medium | |
| WO2021248564A1 (en) | Panoramic big data application monitoring and control system | |
| KR102171384B1 (en) | Object recognition system and method using image correction filter | |
| CN114511592B (en) | Personnel track tracking method and system based on RGBD camera and BIM system | |
| CN113743380B (en) | Active tracking method based on video image dynamic monitoring | |
| CN111144260A (en) | Detection method, device and system of crossing gate | |
| CN116862832A (en) | Three-dimensional live-action model-based operator positioning method | |
| CN112767452B (en) | Active sensing method and system for camera | |
| CN114511592A (en) | Personnel trajectory tracking method and system based on RGBD camera and BIM system | |
| CN111754713B (en) | Video monitoring method, device and system | |
| Chowdhury et al. | Human surveillance system for security application | |
| CN115052109B (en) | Target positioning method and system based on multiple types of cameras | |
| CN113887384B (en) | Pedestrian track analysis method, device, equipment and medium based on multi-track fusion | |
| CN114979567B (en) | Object and region interaction method and system applied to video intelligent monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |