CN111462200A - Cross-video pedestrian positioning and tracking method, system and equipment - Google Patents
Cross-video pedestrian positioning and tracking method, system and equipment Download PDFInfo
- Publication number
- CN111462200A CN111462200A CN202010259428.3A CN202010259428A CN111462200A CN 111462200 A CN111462200 A CN 111462200A CN 202010259428 A CN202010259428 A CN 202010259428A CN 111462200 A CN111462200 A CN 111462200A
- Authority
- CN
- China
- Prior art keywords
- video
- pedestrian
- geographic
- monitoring
- tracked
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 95
- 238000012544 monitoring process Methods 0.000 claims abstract description 98
- 238000001514 detection method Methods 0.000 claims abstract description 31
- 238000007476 Maximum Likelihood Methods 0.000 claims abstract description 24
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 238000004458 analytical method Methods 0.000 claims abstract description 15
- 230000015654 memory Effects 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000003708 edge detection Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 230000000452 restraining effect Effects 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 abstract description 4
- 239000011159 matrix material Substances 0.000 description 14
- 238000010586 diagram Methods 0.000 description 11
- 238000003384 imaging method Methods 0.000 description 10
- 238000003860 storage Methods 0.000 description 7
- 230000009466 transformation Effects 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000004590 computer program Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 230000001629 suppression Effects 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000011541 reaction mixture Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30241—Trajectory
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Remote Sensing (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The application relates to a cross-video pedestrian positioning and tracking method, a cross-video pedestrian positioning and tracking system and electronic equipment. The method comprises the following steps: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video; calculating the pedestrian detection position of the monitoring video to obtain the geographic position of the pedestrian; and combining the pedestrian position information detected by a plurality of adjacent videos, and performing pedestrian cross-video re-recognition analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrian. The pedestrian geographic position information is introduced based on video geographic calibration, maximum likelihood estimation is carried out when the cross-video pedestrian re-identifies, the difficulty and the system complexity of the visual personnel re-identifying track tracking algorithm are reduced, and the method has better application value in a multi-path cross-video system scene.
Description
Technical Field
The application belongs to the technical field of pedestrian positioning and tracking, and particularly relates to a cross-video pedestrian positioning and tracking method, a cross-video pedestrian positioning and tracking system and electronic equipment.
Background
The existing pedestrian positioning method mainly adopts an environment measurement positioning method and a positioning method based on a camera. The environment measurement positioning method refers to positioning or ranging of pedestrians outdoors. Can pass through laser rangefinder or ultrasonic ranging in military field, the precision reaches centimetre level. Although such methods are highly accurate and provide geographic location information for pedestrians, continuous, wide-range, multi-pedestrian determination is difficult.
The positioning method based on the camera comprises binocular positioning and monocular positioning, wherein the binocular positioning is applied to the field of accurate positioning of the robot, and is not suitable for a camera monitoring system in a public area through an S L AM (massive localization and mapping) method for constructing a three-dimensional space, and the method is used for instant positioning and map construction.
The geographic information construction method mainly judges the geographic position information of the image based on the image content. It is generally necessary to train a pre-existing database of images with geographic location labels to obtain a classifier or to find images from it that are similar to the query image. Therefore, image-based geographic location identification requires first extracting visual features from images for comparing similarities between different images. However, in the conventional monitoring scenario, the prior image data cannot be acquired in advance.
The positioning method based on the camera does not belong to the measurement field, so that the pedestrian geographic information can be acquired only by introducing an image geographic information construction method. In the field of cross-video tracking, a method of introducing SIFT parameters is mainly adopted for the same pedestrian tracking, namely, the details of the pedestrian to be tracked are obtained as much as possible and stored in a database, and when the same pedestrian appears, the pedestrian matched with the database is re-identified. In the field of re-identification, a common cross-video tracking method comprises template matching, searching an image area to be matched by using a given template, and obtaining a matching result according to the calculated matching degree. However, the pedestrian matching method needs to give a template in advance, the requirement on the area to be matched to be searched is strict, and due to the existence of matching and searching, the calculation efficiency of continuous videos is low, the time consumption is long, and the method cannot be completely applied to a multi-video system.
In addition, the existing pedestrian tracking system mainly adopts methods of searching pedestrians and positioning pedestrians, and when cross-video tracking is carried out under multi-video linkage, methods of more pedestrian characteristic parameters are introduced for tracking. However, the existing pedestrian tracking system is difficult to acquire the continuous space-time trajectory of the pedestrian because the position information of the pedestrian cannot be acquired.
Disclosure of Invention
The application provides a cross-video pedestrian positioning and tracking method, a cross-video pedestrian positioning and tracking system and electronic equipment, and aims to solve at least one of the technical problems in the prior art to a certain extent.
In order to solve the above problems, the present application provides the following technical solutions:
a cross-video pedestrian positioning and tracking method comprises the following steps:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step a, the performing reference point topology matching and video geographic registration on the surveillance video according to object identification specifically includes:
adopting an object recognition algorithm to perform object recognition classification on the monitoring video to obtain reference points in the monitoring video;
matching the reference point with an object in the object geographic database to obtain geographic position information of the control point with the same name in the monitoring video;
and performing geographic registration on the homonymous control points of the ground area in the monitoring video by adopting a world geographic coordinate system conversion method, so that the monitoring video has geographic position information.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step a, the performing reference point topology matching and video geographic registration on the surveillance video according to object identification further includes:
carrying out image preprocessing on the monitoring video to obtain a video image subjected to fisheye calibration;
and intercepting a certain frame of two-dimensional image in the preprocessed video image, and performing edge extraction on the two-dimensional image by adopting an edge detection and watershed segmentation method to obtain a ground area with GIS information in the two-dimensional image.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step b, the calculating the pedestrian detection position of the surveillance video specifically comprises:
and detecting a moving object in the monitoring video by adopting a frame difference method, positioning the position of the pedestrian to be tracked by combining a human head detector, and acquiring the head information of the pedestrian to be tracked.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the human head detector adopts a pedestrian detection method based on a convolutional neural network CNN, the convolutional neural network comprises an input layer, convolutional layers, a pooling layer, a full-connection layer and an output layer, the convolutional layers and the pooling layer are compounded to process input data, and mapping between the convolutional layers and the pooling layer and an output target is carried out through the connection layer.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in step b, the step of obtaining the geographic position of the pedestrian further comprises:
and carrying out movement detection on the pedestrian to be tracked based on the head information to obtain underfoot pixel points of the pedestrian to be tracked, wherein the underfoot pixel points are the geographical position information of the pedestrian to be tracked.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in step b, the step of obtaining the geographic position of the pedestrian further comprises:
and calibrating the geographical position information of the pedestrian to be tracked by a method of restraining a camera.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in step c, the performing pedestrian cross-video re-recognition analysis on the multiple paths of monitoring videos by using the maximum likelihood estimation method further includes:
when shooting scene movement exists in the multi-path monitoring videos, triggering geographic area overlapping judgment; the geographic area overlap determination specifically includes:
positioning the geographic information of the shooting scenes of the multiple paths of monitoring videos, and dividing the monitoring areas of all the camera devices according to the overlapped geographic position areas; obtaining the continuous track of the pedestrian to be tracked by connecting the geographic information space coordinates of the pedestrian to be tracked in the continuous frames, and triggering the next camera to track the pedestrian track when the geographic information space coordinates of the pedestrian to be tracked exceed the monitoring area of the current camera and move to the monitoring area of the next camera.
Another technical scheme adopted by the embodiment of the application is as follows: a cross-video pedestrian location tracking system, comprising:
a video geographic registration module: the system is used for constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
the cross-video pedestrian positioning and tracking module: the pedestrian detection position calculation is carried out on the monitoring video to obtain the geographic position of the pedestrian to be tracked;
the multi-video track tracking module: the method is used for carrying out pedestrian cross-video re-identification analysis on the multi-path monitoring video by combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos and adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
The embodiment of the application adopts another technical scheme that: an electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the one processor to cause the at least one processor to perform the following operations of the cross-video pedestrian location tracking method described above:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
Compared with the prior art, the embodiment of the application has the advantages that: the cross-video pedestrian positioning and tracking method, system and electronic equipment acquire reference object position information through object identification to conduct geographic registration of monitoring videos, acquire pedestrian geographic position information through pedestrian detection and acquire moving space-time tracks of the pedestrians, are simple in operation, have geographic position information, and have better application value in a multi-channel video system scene. The pedestrian geographic position information is introduced based on video geographic calibration, maximum likelihood estimation is carried out when the cross-video pedestrian re-identifies, the difficulty and the system complexity of the visual personnel re-identifying track tracking algorithm are reduced, and the method has better application value in a multi-path cross-video system scene.
Drawings
FIG. 1 is a flow chart of a cross-video pedestrian location tracking method according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a surveillance video geographic registration algorithm according to an embodiment of the present application;
FIG. 3(a) is a video image before preprocessing, and FIG. 3(b) is a video image after preprocessing;
FIG. 4 is a schematic view of an imaging model of a camera;
fig. 5(a) and (b) are schematic diagrams of spatial relationship between pixel coordinates and world geographic coordinates, where fig. 5(a) is the pixel coordinates and fig. 5(b) is the world geographic coordinates;
FIG. 6 is a schematic diagram of a pedestrian detection algorithm in accordance with an embodiment of the present application;
FIG. 7 is a flow chart of a frame differencing method according to an embodiment of the application;
FIG. 8 is an exemplary diagram of maximum likelihood estimation;
FIG. 9 is a flowchart of a cross-video pedestrian tracking algorithm based on a geographic area overlap determination according to an embodiment of the present application;
FIG. 10 is a schematic structural diagram of a cross-video pedestrian location tracking system according to an embodiment of the present application;
fig. 11 is a schematic structural diagram of a hardware device of a cross-video pedestrian location and tracking method according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Please refer to fig. 1, which is a flowchart illustrating a cross-video pedestrian location tracking method according to an embodiment of the present application. The cross-video pedestrian positioning and tracking method comprises the following steps:
step 100: acquiring a monitoring video of a pedestrian to be tracked;
step 200: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on the monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
in this step, the object geographic coordinate database includes a third-party geographic information base such as hundredths or existing BIM file data (data for construction engineering including building appearance and geographic position). In order to obtain more accurate GPS position information, a standard WGS84 coordinate system mode is adopted for reference point calibration, BIM (Building information model) information of an identified object corresponds to a coordinate point of a public area WGS84 coordinate system and serves as a reference point, and therefore the accuracy of pedestrian tracking space calculation is improved. The object geographic database is applied to correspondence of a world coordinate system and an image coordinate system after conversion, and is used for identifying an object by utilizing the existing object geographic information and an image, registering pixel points in a monitoring video with actual coordinates, and determining the spatial position coordinates of the monitoring video.
Further, please refer to fig. 2, which is a schematic diagram of a geographic registration algorithm of a surveillance video according to an embodiment of the present application. The geographic registration algorithm for the surveillance video in the embodiment of the application comprises the following steps:
step 210: adopting an object recognition algorithm to perform object recognition classification on the monitoring video to obtain reference points in the monitoring video, and simultaneously performing image preprocessing on the monitoring video to obtain a video image subjected to fisheye calibration;
in step 210, the object recognition algorithm includes methods such as RCNN and YO L O, and performs object recognition and classification on the surveillance video to obtain a reference point in the surveillance video, and performs fuzzy matching on the reference point and an object in the object geographic database.
In practical situations, due to the influence of factors such as the manufacturing level and the assembly level of an imaging original, an edge distortion condition exists in a video shooting process of a camera, and the edge distortion of a lens with a wider angle is more obvious, so that the nonlinear distortion of a monitoring video is caused, and therefore the nonlinear distortion of the camera needs to be corrected, and the calculation accuracy is improved. The method comprises the steps of preprocessing a monitoring video through a chessboard correction method, calculating internal parameters and correction coefficients of fisheyes, correcting and cutting the monitoring video, removing part of edge parts, and obtaining a video image after fisheye correction, wherein the preprocessed video image can reduce errors caused by nonlinear distortion during world coordinate system conversion. Specifically, as shown in fig. 3(a) and (b), fig. 3(a) is a video image before preprocessing, and fig. 3(b) is a video image after preprocessing.
Specifically, the nonlinear distortion is generally geometric distortion, which causes a certain offset between the coordinates of the pixel point and the coordinates of the ideal pixel point, and can be expressed as:
in the formula (1), (u, v) is an ideal pixel coordinate, and (u ', v') is a pixel coordinate affected by distortion. Non-linear distortionu、vThe expression of (c) can be written as:
in the formula (2), the reaction mixture is,u、vthe second and third terms being due to inaccuracies in the imaging original of the camera, parameter p1、p2、k1、k2、s1、s2Is a nonlinear distortion parameter. And restoring the image distortion by calculating the numerical value of the nonlinear distortion parameter.
Step 220: geographic registration is carried out on the identified reference point and an object in an object geographic database, and geographic position information of the control point with the same name in the monitoring video is obtained;
step 230: intercepting a certain frame of two-dimensional image in the preprocessed video image, and performing edge extraction on the two-dimensional image by adopting an edge detection and watershed segmentation method to obtain a ground area with GIS information in the two-dimensional image;
in step 230, by using the special spatial structure relationship of the surveillance video, the area below the video is the ground portion, and the vertical object can block the ground. Therefore, the horizontal structure and the vertical structure of the image are calculated by the edge detection technology and the watershed algorithm.
Step 240: geographic registration is carried out on the homonymous control points of the ground area in the monitoring video by adopting a world geographic coordinate system conversion method, so that the monitoring video has geographic position information;
in step 240, the coordinate system is converted and calculated according to the world geographic coordinate systemAnd shifting the matrix, and respectively performing geographic coordinate matching on each pixel point of the ground area in the monitoring video through methods such as image stretching, filling, cutting and the like to obtain a plane with geographic information in the monitoring video. The actual geographic position of the identified object and the relative position of the observed object are obtained by comparing the object geographic database. Since the matrix calculation generally performs the conversion between the pixel coordinates of one plane and the world coordinates through four points, the result obtained by the world coordinate system conversion matrix calculation is an estimated value and is not an accurate value. In order to control the error of the estimated value, the image is subjected to constraint calculation under the condition that more control points with the same name exist. In a plurality of three-dimensional space points Xw＝(Xw，YW，ZW)TProjecting to an imaging plane to obtain a corresponding two-dimensional plane point m ═ u, v)TThen, a plurality of three-dimensional space points X are obtained through a plurality of times of matrix calculation of the triangular areaw＝(Xw，YW，ZW)T. By introducing a method of calculating and averaging a plurality of homonymous control points for a plurality of times, the calculation error caused by overlarge deformation amount can be reduced.
The camera imaging model is shown in fig. 4. A three-dimensional space point X in the objective worldw＝(Xw，YW，ZW)TProjecting to an imaging plane to obtain a corresponding two-dimensional plane point m ═ u, v)TAnd the method is realized according to coordinate transformation between different coordinate systems. Specifically, the coordinate transformation formula between the world coordinate system and the camera coordinate system is as follows:
in the formula (3), Xc＝(Xc，Yc，Zc)TRepresents point XWAnd (3) three-dimensional coordinates in a camera coordinate system, wherein R is a rotation transformation matrix of 3 × 3, t is a translation transformation matrix of 3 × 1, and R and t respectively represent the relative attitude and position between the world coordinate system and the camera coordinate system.
The coordinate transformation method between the camera coordinate system and the image coordinate system is to project a three-dimensional space point P (X, Y, Z) in the camera coordinate system to the imaging plane to obtain a corresponding two-dimensional plane point P (X, Y), and the relationship between X, Y and X, Y can be expressed as:
in the formula (4), f is the focal length of the camera. The geographic position corresponding to the area shot by the camera can be calculated by the known world geographic coordinate plane point P (X, Y, Z) with geographic position information and the two-dimensional coordinate point P (X, Y) of the imaging plane. Specifically, as shown in fig. 4, a schematic view of an imaging model of a camera is shown.
According to an image coordinate system o 'xy and a pixel coordinate system o'xyThe coordinate relationship of (a) can be given by:
again, the spatial geometry can be found:
the distance of the identified object or pixel region in the vertical direction on the ground can be obtained by combining the formula as follows:
therefore, the height of the recognized object can be judged according to the Y value, and when the Y value is smaller than the threshold value of the set ground (the accurate value can be set according to the actual situation), the ground area can be judged, and the ground coordinates can be determined. By calculating Xc＝(Xc，Yc，Zc)TAnd calculating an offset matrix with the pixel point p (x, y), and deforming a ground area of a certain block in the picture into an actual ground matrix matched with the object geographic database in a matrix offset and filling mode to realize the correspondence of the pixel coordinate and the geographic coordinate. FIG. 5(a) and (b) are imagesThe spatial relationship between the pixel coordinates and the world geographic coordinates is illustrated schematically, where fig. 5(a) is the pixel coordinates and fig. 5(b) is the world geographic coordinates.
In the world coordinate system conversion, because the image can have the skew error when the formation of image, can lead to passing through transfer matrix world coordinate system error bigger, consequently, for the suppression error, this application provides the method of multiple suppression error and carries out position calibration to pedestrian's pixel. After the object is identified, the bottom center point of the object identification frame is used as the geographic position of the object and is used as a reference point for calculation; first, the relationship between a two-dimensional plane point (x, y) in an image and its corresponding point (u, v) in a pixel coordinate system is represented by:
the above equation can be expressed in a coordinate transformation form as:
in the formula (9), dx，dyRepresenting the physical distance of the pixel in the u and v axes and directions of the image capture (u)0，v0) Is the coordinates of the camera principal point in the pixel coordinate system. By simultaneous obtaining:
in the formula (10), fu，fvRespectively, represent the focal length in units of pixel width and height. Matrix array
The parameters in (1) are called camera intrinsic parameters, and are only influenced by the internal structure and imaging characteristics of the camera. The parameters of matrix R and matrix t are referred to as the camera-out parameters. P ═ K [ R, t ═]Referred to as a perspective projection matrix. Therefore, the conversion between the plane point and the camera point is realized, the distance between any point (x, y) and the known point (u, v) can be calculated, and further, the distance between any point (x, y) and the known point (u, v) can be calculatedThe GPS information of the point (x, y) is calculated.
Step 300: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
please refer to fig. 6, which is a schematic diagram of a pedestrian detection algorithm according to an embodiment of the present application. The pedestrian detection algorithm of the embodiment of the application comprises the following steps:
step 310: carrying out human head detection on the monitoring video to acquire head information of a pedestrian to be tracked;
in step 310, human head detection is a method for rapidly identifying a pedestrian head model, and is suitable for multiple paths of monitoring videos. In order to improve the timeliness of pedestrian detection, the method adopts a frame difference method to detect a moving object in a monitoring video, and is combined with a human head detector to position the position of a pedestrian. The human head detector adopts a pedestrian detection method based on a convolutional neural network CNN, the convolutional neural network is composed of an input layer, convolutional layers, pooling layers, full-connection layers and output layers, the convolutional layers and the pooling layers are compounded to process input data, and mapping between the convolutional layers and the pooling layers and output pedestrians is realized on the connection layers. Fig. 7 shows a flow chart of the frame difference method, which specifically includes: the images of the n +1 th frame, the n-1 th frame and the n-1 th frame in the video sequence are respectively recorded as fn +1, fn and fn-1, the gray values of the corresponding pixel points of the three frames of images are respectively recorded as fn +1(x, y), fn (x, y) and fn-1(x, y), differential images Dn +1 and Dn are respectively obtained, the differential images Dn +1 and Dn are operated, then threshold processing and connectivity analysis are carried out, and finally a moving object is detected.
Step 320: performing movement detection (pedestrian position calculation) on the pedestrian to be tracked based on the pedestrian head detection result to obtain underfoot pixel points of the pedestrian to be tracked, namely the geographic position information of the pedestrian to be tracked;
in step 320, because the pedestrian is a moving object, according to the characteristics of the video, the lowermost region of the moving object is a position where the foot stands, that is, the positioning coordinate point of the pedestrian to be tracked is a foot region, and the head coordinate point cannot be used as a spatial coordinate pedestrian calculated by the geographic position, so that the method of combining head detection and movement detection is used for quickly finding the underfoot pixel point corresponding to the moving region of the pedestrian to be tracked.
Step 330: calibrating the geographical position information of the pedestrian to be tracked;
in step 330, the method calibrates the geographic position information of the pedestrian to be tracked by inhibiting the camera, so that the uncertain error caused by the blur when the pedestrian moves is reduced, and the positioning accuracy is improved.
Step 400: combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-recognition analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain continuous space-time tracks of the pedestrians to be tracked;
in step 400, when a pedestrian moves across videos, although the pedestrian has geographic position information of the pedestrian to be tracked, whether the pedestrian is the same pedestrian cannot be determined, the method determines whether the pedestrian is the same pedestrian through probability calculation by performing maximum likelihood estimation on the movement track of the same pedestrian to be tracked in a plurality of videos. The method specifically comprises the following steps: given a probability distribution D, assume its probability density function (continuous distribution) or probability aggregation function (discrete distribution) to be fDAnd a distribution parameter θ, from which a sample x having n values can be extracted1，x2，...，xnThen, θ is estimated. By using fDThe probability is calculated as follows:
P＝(x1，x2，x3...，xn)＝fD(x1，x2，x3...，xn|θ) (11)
the maximum likelihood estimate finds the most likely value for theta (i.e., finding one value that maximizes the "likelihood" of this sample among all possible theta values). To mathematically implement the maximum likelihood estimation method, first the likelihood is defined:
lik(θ)＝fD(x1，x2，x3...，xn|θ) (12)
and this function is maximized over all values of theta. This value that maximizes the likelihood is the maximum likelihood estimate of θ. And comparing the tracks obtained from the two videos to judge the maximum likelihood, and judging whether the videos are the same pedestrian or not after setting a threshold value. Specifically, as shown in fig. 8, it is an exemplary diagram of maximum likelihood estimation.
When the pedestrian to be tracked moves across videos, the shooting scene may move, and at this time, the geographic area overlap determination is triggered, specifically as shown in fig. 9, which is a flowchart of a cross-video pedestrian tracking algorithm based on the geographic area overlap determination according to the embodiment of the present application. And the geographic area overlapping judgment is to position the geographic information of the shooting scenes of the multiple paths of monitoring videos and then divide the respective tracked monitoring areas according to the overlapped geographic position areas. The continuous track of the pedestrian to be tracked is obtained by connecting the geographic information space coordinates of the pedestrian to be tracked in the continuous frames, and when the geographic information space coordinates of the pedestrian to be tracked exceed the monitoring area of the current camera and move to the monitoring area of the next camera, the next camera is triggered to track the pedestrian track.
Please refer to fig. 10, which is a schematic structural diagram of a cross-video pedestrian positioning and tracking system according to an embodiment of the present application. The cross-video pedestrian positioning and tracking system of the embodiment of the application comprises:
a video geographic registration module: the system is used for constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
the cross-video pedestrian positioning and tracking module: the system is used for calculating the pedestrian detection position of the monitoring video and acquiring the geographic position of the pedestrian to be tracked;
the multi-video track tracking module: the method is used for combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
Fig. 11 is a schematic structural diagram of a hardware device of a cross-video pedestrian location and tracking method according to an embodiment of the present application. As shown in fig. 11, the device includes one or more processors and memory. Taking a processor as an example, the apparatus may further include: an input system and an output system.
The processor, memory, input system, and output system may be connected by a bus or other means, as exemplified by the bus connection in fig. 11.
The memory, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and modules. The processor executes various functional applications and data processing of the electronic device, i.e., implements the processing method of the above-described method embodiment, by executing the non-transitory software program, instructions and modules stored in the memory.
The memory may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data and the like. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory located remotely from the processor, and these remote memories may be connected to the processing system over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input system may receive input numeric or character information and generate a signal input. The output system may include a display device such as a display screen.
The one or more modules are stored in the memory and, when executed by the one or more processors, perform the following for any of the above method embodiments:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
The product can execute the method provided by the embodiment of the application, and has the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the methods provided in the embodiments of the present application.
Embodiments of the present application provide a non-transitory (non-volatile) computer storage medium having stored thereon computer-executable instructions that may perform the following operations:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
Embodiments of the present application provide a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions that, when executed by a computer, cause the computer to perform the following:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
The cross-video pedestrian positioning and tracking method, system and electronic equipment acquire reference object position information through object identification to conduct geographic registration of monitoring videos, acquire pedestrian geographic position information through pedestrian detection and acquire moving space-time tracks of the pedestrians, are simple in operation, have geographic position information, and have better application value in a multi-channel video system scene. Meanwhile, pedestrian geographic position information is introduced based on video geographic calibration, maximum likelihood estimation is carried out when the cross-video pedestrian is re-identified, the difficulty of the visual staff re-identifying the track tracking algorithm and the system complexity are reduced, and the method has better application value in a multi-path cross-video system scene.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A cross-video pedestrian positioning and tracking method is characterized by comprising the following steps:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
2. The cross-video pedestrian positioning and tracking method according to claim 1, wherein in the step a, the performing reference point topology matching and video geographic registration on the surveillance video according to object recognition specifically comprises:
adopting an object recognition algorithm to perform object recognition classification on the monitoring video to obtain reference points in the monitoring video;
matching the reference point with an object in the object geographic database to obtain geographic position information of the control point with the same name in the monitoring video;
and performing geographic registration on the homonymous control points of the ground area in the monitoring video by adopting a world geographic coordinate system conversion method, so that the monitoring video has geographic position information.
3. The cross-video pedestrian positioning and tracking method according to claim 2, wherein in the step a, the performing reference point topology matching and video geographic registration on the surveillance video according to object recognition further comprises:
carrying out image preprocessing on the monitoring video to obtain a video image subjected to fisheye calibration;
and intercepting a certain frame of two-dimensional image in the preprocessed video image, and performing edge extraction on the two-dimensional image by adopting an edge detection and watershed segmentation method to obtain a ground area with GIS information in the two-dimensional image.
4. The cross-video pedestrian location tracking method according to any one of claims 1 to 3, wherein in the step b, the calculating the pedestrian detection position of the surveillance video specifically comprises:
and detecting a moving object in the monitoring video by adopting a frame difference method, positioning the position of the pedestrian to be tracked by combining a human head detector, and acquiring the head information of the pedestrian to be tracked.
5. The cross-video pedestrian positioning and tracking method according to claim 4, wherein the human head detector adopts a pedestrian detection method based on a Convolutional Neural Network (CNN), the convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, a full-link layer and an output layer, the plurality of convolutional layers and pooling layers are compounded to process input data, and mapping between the input data and an output target is performed through the link layers.
6. The cross-video pedestrian location tracking method according to claim 4, wherein in the step b, the obtaining the geographic position of the pedestrian further comprises:
and carrying out movement detection on the pedestrian to be tracked based on the head information to obtain underfoot pixel points of the pedestrian to be tracked, wherein the underfoot pixel points are the geographical position information of the pedestrian to be tracked.
7. The cross-video pedestrian location tracking method according to claim 6, wherein in the step b, the obtaining the pedestrian geographic position further comprises:
and calibrating the geographical position information of the pedestrian to be tracked by a method of restraining a camera.
8. The cross-video pedestrian location tracking method according to claim 7, wherein in the step c, the performing pedestrian cross-video re-recognition analysis on the multiple surveillance videos by using the maximum likelihood estimation method further comprises:
when shooting scene movement exists in the multi-path monitoring videos, triggering geographic area overlapping judgment; the geographic area overlap determination specifically includes:
positioning the geographic information of the shooting scenes of the multiple paths of monitoring videos, and dividing the monitoring areas of all the camera devices according to the overlapped geographic position areas; obtaining the continuous track of the pedestrian to be tracked by connecting the geographic information space coordinates of the pedestrian to be tracked in the continuous frames, and triggering the next camera to track the pedestrian track when the geographic information space coordinates of the pedestrian to be tracked exceed the monitoring area of the current camera and move to the monitoring area of the next camera.
9. A cross-video pedestrian location tracking system, comprising:
a video geographic registration module: the system is used for constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
the cross-video pedestrian positioning and tracking module: the pedestrian detection position calculation is carried out on the monitoring video to obtain the geographic position of the pedestrian to be tracked;
the multi-video track tracking module: the method is used for carrying out pedestrian cross-video re-identification analysis on the multi-path monitoring video by combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos and adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
10. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the following operations of the cross-video pedestrian location tracking method of any of claims 1 to 8 above:
step a: constructing an object geographic coordinate database, carrying out reference point topological matching and video geographic registration on a monitoring video according to object identification, and determining geographic coordinates of pixel points of the monitoring video;
step b: calculating the pedestrian detection position of the monitoring video to acquire the geographic position of the pedestrian to be tracked;
step c: and combining the geographic positions of the pedestrians to be tracked detected by the adjacent videos, and performing pedestrian cross-video re-identification analysis on the multi-path monitoring videos by adopting a maximum likelihood estimation method to obtain the continuous space-time trajectory of the pedestrians to be tracked.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010259428.3A CN111462200B (en) | 2020-04-03 | 2020-04-03 | Cross-video pedestrian positioning and tracking method, system and equipment |
| PCT/CN2020/085081 WO2021196294A1 (en) | 2020-04-03 | 2020-04-16 | Cross-video person location tracking method and system, and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010259428.3A CN111462200B (en) | 2020-04-03 | 2020-04-03 | Cross-video pedestrian positioning and tracking method, system and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111462200A true CN111462200A (en) | 2020-07-28 |
| CN111462200B CN111462200B (en) | 2023-09-19 |
Family
ID=71680274
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010259428.3A Active CN111462200B (en) | 2020-04-03 | 2020-04-03 | Cross-video pedestrian positioning and tracking method, system and equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111462200B (en) |
| WO (1) | WO2021196294A1 (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070003A (en) * | 2020-09-07 | 2020-12-11 | 深延科技(北京)有限公司 | Personnel tracking method and system based on deep learning |
| CN112163537A (en) * | 2020-09-30 | 2021-01-01 | 中国科学院深圳先进技术研究院 | Pedestrian abnormal behavior detection method, system, terminal and storage medium |
| CN112184814A (en) * | 2020-09-24 | 2021-01-05 | 天津锋物科技有限公司 | Positioning method and positioning system |
| CN112733657A (en) * | 2020-12-31 | 2021-04-30 | 罗普特科技集团股份有限公司 | Cross-border tracking detection method and system based on standard address and POI information point |
| CN112766210A (en) * | 2021-01-29 | 2021-05-07 | 苏州思萃融合基建技术研究所有限公司 | Safety monitoring method and device for building construction and storage medium |
| CN113190711A (en) * | 2021-03-26 | 2021-07-30 | 南京财经大学 | Video dynamic object trajectory space-time retrieval method and system in geographic scene |
| CN113435329A (en) * | 2021-06-25 | 2021-09-24 | 湖南大学 | Unsupervised pedestrian re-identification method based on video track feature association learning |
| CN113837023A (en) * | 2021-09-02 | 2021-12-24 | 北京新橙智慧科技发展有限公司 | Cross-camera pedestrian automatic tracking method |
| WO2022067606A1 (en) * | 2020-09-30 | 2022-04-07 | 中国科学院深圳先进技术研究院 | Method and system for detecting abnormal behavior of pedestrian, and terminal and storage medium |
| CN114842028A (en) * | 2022-05-07 | 2022-08-02 | 深圳先进技术研究院 | Cross-video target tracking method, system, electronic equipment and storage medium |
| CN117237418A (en) * | 2023-11-15 | 2023-12-15 | 成都航空职业技术学院 | Moving object detection method and system based on deep learning |
| CN113627497B (en) * | 2021-07-27 | 2024-03-12 | 武汉大学 | Space-time constraint-based cross-camera pedestrian track matching method |
Families Citing this family (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113947742B (en) * | 2021-10-19 | 2024-09-06 | 成都佳发安泰教育科技股份有限公司 | Human track tracking method and device based on face recognition |
| CN114240997B (en) * | 2021-11-16 | 2023-07-28 | 南京云牛智能科技有限公司 | Intelligent building online trans-camera multi-target tracking method |
| CN114040168A (en) * | 2021-11-16 | 2022-02-11 | 西安热工研究院有限公司 | A intelligent electric power network monitoring mechanism for thermal power plant |
| CN114550041B (en) * | 2022-02-18 | 2024-03-29 | 中国科学技术大学 | Multi-target labeling method for shooting video by multiple cameras |
| CN114785983B (en) * | 2022-03-07 | 2024-10-18 | 复旦大学 | Target cross-domain tracking method based on node communication and equipment cooperation |
| CN114972421B (en) * | 2022-04-27 | 2024-10-18 | 中南大学 | Workshop material identification tracking and positioning method and system |
| CN115527162B (en) * | 2022-05-18 | 2023-07-18 | 湖北大学 | Multi-pedestrian re-identification method and system based on three-dimensional space |
| CN115033960B (en) * | 2022-06-09 | 2023-04-07 | 中国公路工程咨询集团有限公司 | Automatic fusion method and device of BIM (building information modeling) model and GIS (geographic information system) |
| CN115035163A (en) * | 2022-07-04 | 2022-09-09 | 上海易同科技股份有限公司 | Target tracking method, device, equipment and storage medium based on Bluetooth positioning |
| CN114862973B (en) * | 2022-07-11 | 2022-09-16 | 中铁电气化局集团有限公司 | Space positioning method, device and equipment based on fixed point location and storage medium |
| CN115731287B (en) * | 2022-09-07 | 2023-06-23 | 滁州学院 | Moving target retrieval method based on aggregation and topological space |
| CN115578756B (en) * | 2022-11-08 | 2023-04-14 | 杭州昊恒科技有限公司 | Personnel fine management method and system based on precise positioning and video linkage |
| CN115457449B (en) * | 2022-11-11 | 2023-03-24 | 深圳市马博士网络科技有限公司 | Early warning system based on AI video analysis and monitoring security protection |
| CN115856980B (en) * | 2022-11-21 | 2023-08-01 | 中铁科学技术开发有限公司 | Marshalling station operator monitoring method and system |
| CN115808170B (en) * | 2023-02-09 | 2023-06-06 | 宝略科技(浙江)有限公司 | Indoor real-time positioning method integrating Bluetooth and video analysis |
| CN115979250B (en) * | 2023-03-20 | 2023-06-09 | 山东上水环境科技集团有限公司 | Positioning method based on UWB module, semantic map and visual information |
| CN116189116B (en) * | 2023-04-24 | 2024-02-23 | 江西方兴科技股份有限公司 | Traffic state sensing method and system |
| CN116631596B (en) * | 2023-07-24 | 2024-01-02 | 深圳市微能信息科技有限公司 | Monitoring management system and method for working time of radiological personnel |
| CN116740878B (en) * | 2023-08-15 | 2023-12-26 | 广东威恒输变电工程有限公司 | Positioning early warning method for bidirectional drawing of global area coordinated by multiple cameras |
| CN117185064B (en) * | 2023-08-18 | 2024-03-05 | 山东五棵松电气科技有限公司 | Intelligent community management system, method, computer equipment and storage medium |
| CN117173215B (en) * | 2023-09-04 | 2024-08-20 | 东南大学 | Inland navigation ship whole-course track identification method and system crossing cameras |
| CN117058331B (en) * | 2023-10-13 | 2023-12-19 | 山东建筑大学 | Indoor personnel three-dimensional track reconstruction method and system based on single monitoring camera |
| CN118038008B (en) * | 2024-04-15 | 2024-07-12 | 武汉人云智物科技有限公司 | Hydropower plant personnel positioning method and system based on ptz multi-camera linkage |
| CN118429947B (en) * | 2024-07-02 | 2024-09-03 | 汶上义桥煤矿有限责任公司 | Intelligent early warning method and system for preventing misoperation of monorail crane |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101548639B1 (en) * | 2014-12-10 | 2015-09-01 | 한국건설기술연구원 | Apparatus for tracking the objects in surveillance camera system and method thereof |
| CN107153824A (en) * | 2017-05-22 | 2017-09-12 | 中国人民解放军国防科学技术大学 | Across video pedestrian recognition methods again based on figure cluster |
| US20180075593A1 (en) * | 2016-09-15 | 2018-03-15 | Qualcomm Incorporated | Automatic scene calibration method for video analytics |
| CN109461132A (en) * | 2018-10-31 | 2019-03-12 | 中国人民解放军国防科技大学 | SAR image automatic registration method based on feature point geometric topological relation |
| WO2019145018A1 (en) * | 2018-01-23 | 2019-08-01 | Siemens Aktiengesellschaft | System, device and method for detecting abnormal traffic events in a geographical location |
| CN110147471A (en) * | 2019-04-04 | 2019-08-20 | 平安科技（深圳）有限公司 | Trace tracking method, device, computer equipment and storage medium based on video |
| CN110375739A (en) * | 2019-06-26 | 2019-10-25 | 中国科学院深圳先进技术研究院 | A kind of mobile terminal vision fusion and positioning method, system and electronic equipment |
| CN110706259A (en) * | 2019-10-12 | 2020-01-17 | 四川航天神坤科技有限公司 | Space constraint-based cross-shot tracking method and device for suspicious people |
| CN110717414A (en) * | 2019-09-24 | 2020-01-21 | 青岛海信网络科技股份有限公司 | Target detection tracking method, device and equipment |
| CN110765903A (en) * | 2019-10-10 | 2020-02-07 | 浙江大华技术股份有限公司 | Pedestrian re-identification method and device and storage medium |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105472332B (en) * | 2015-12-01 | 2018-07-20 | 广东弘栎电子设备有限公司 | Analysis method based on location technology and video technique and its analysis system |
| CN105913037A (en) * | 2016-04-26 | 2016-08-31 | 广东技术师范学院 | Face identification and radio frequency identification based monitoring and tracking system |
| WO2018087545A1 (en) * | 2016-11-08 | 2018-05-17 | Staffordshire University | Object location technique |
| CN107547865A (en) * | 2017-07-06 | 2018-01-05 | 王连圭 | Trans-regional human body video frequency object tracking intelligent control method |
| WO2020055767A1 (en) * | 2018-09-10 | 2020-03-19 | Mapbox, Inc. | Mapping objects detected in images to geographic positions |
| CN110414441B (en) * | 2019-07-31 | 2022-05-10 | 浙江大学 | Pedestrian track analysis method and system |
-
2020
- 2020-04-03 CN CN202010259428.3A patent/CN111462200B/en active Active
- 2020-04-16 WO PCT/CN2020/085081 patent/WO2021196294A1/en active Application Filing
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101548639B1 (en) * | 2014-12-10 | 2015-09-01 | 한국건설기술연구원 | Apparatus for tracking the objects in surveillance camera system and method thereof |
| US20180075593A1 (en) * | 2016-09-15 | 2018-03-15 | Qualcomm Incorporated | Automatic scene calibration method for video analytics |
| CN107153824A (en) * | 2017-05-22 | 2017-09-12 | 中国人民解放军国防科学技术大学 | Across video pedestrian recognition methods again based on figure cluster |
| WO2019145018A1 (en) * | 2018-01-23 | 2019-08-01 | Siemens Aktiengesellschaft | System, device and method for detecting abnormal traffic events in a geographical location |
| CN109461132A (en) * | 2018-10-31 | 2019-03-12 | 中国人民解放军国防科技大学 | SAR image automatic registration method based on feature point geometric topological relation |
| CN110147471A (en) * | 2019-04-04 | 2019-08-20 | 平安科技（深圳）有限公司 | Trace tracking method, device, computer equipment and storage medium based on video |
| CN110375739A (en) * | 2019-06-26 | 2019-10-25 | 中国科学院深圳先进技术研究院 | A kind of mobile terminal vision fusion and positioning method, system and electronic equipment |
| CN110717414A (en) * | 2019-09-24 | 2020-01-21 | 青岛海信网络科技股份有限公司 | Target detection tracking method, device and equipment |
| CN110765903A (en) * | 2019-10-10 | 2020-02-07 | 浙江大华技术股份有限公司 | Pedestrian re-identification method and device and storage medium |
| CN110706259A (en) * | 2019-10-12 | 2020-01-17 | 四川航天神坤科技有限公司 | Space constraint-based cross-shot tracking method and device for suspicious people |
Non-Patent Citations (2)
| Title |
|---|
| ADITYA VORA ET AL: "FCHD: FAST AND ACCURATE HEAD DETECTION IN CROWDED SCENES", pages 1 - 5 * |
| 卢文涛 等: "基于拓扑结构的地图匹配算法研究", vol. 29, no. 6, pages 73 - 76 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070003A (en) * | 2020-09-07 | 2020-12-11 | 深延科技(北京)有限公司 | Personnel tracking method and system based on deep learning |
| CN112184814A (en) * | 2020-09-24 | 2021-01-05 | 天津锋物科技有限公司 | Positioning method and positioning system |
| CN112184814B (en) * | 2020-09-24 | 2022-09-02 | 天津锋物科技有限公司 | Positioning method and positioning system |
| WO2022067606A1 (en) * | 2020-09-30 | 2022-04-07 | 中国科学院深圳先进技术研究院 | Method and system for detecting abnormal behavior of pedestrian, and terminal and storage medium |
| CN112163537A (en) * | 2020-09-30 | 2021-01-01 | 中国科学院深圳先进技术研究院 | Pedestrian abnormal behavior detection method, system, terminal and storage medium |
| CN112163537B (en) * | 2020-09-30 | 2024-04-26 | 中国科学院深圳先进技术研究院 | Pedestrian abnormal behavior detection method, system, terminal and storage medium |
| CN112733657A (en) * | 2020-12-31 | 2021-04-30 | 罗普特科技集团股份有限公司 | Cross-border tracking detection method and system based on standard address and POI information point |
| CN112766210A (en) * | 2021-01-29 | 2021-05-07 | 苏州思萃融合基建技术研究所有限公司 | Safety monitoring method and device for building construction and storage medium |
| CN113190711A (en) * | 2021-03-26 | 2021-07-30 | 南京财经大学 | Video dynamic object trajectory space-time retrieval method and system in geographic scene |
| CN113435329B (en) * | 2021-06-25 | 2022-06-21 | 湖南大学 | Unsupervised pedestrian re-identification method based on video track feature association learning |
| CN113435329A (en) * | 2021-06-25 | 2021-09-24 | 湖南大学 | Unsupervised pedestrian re-identification method based on video track feature association learning |
| CN113627497B (en) * | 2021-07-27 | 2024-03-12 | 武汉大学 | Space-time constraint-based cross-camera pedestrian track matching method |
| CN113837023A (en) * | 2021-09-02 | 2021-12-24 | 北京新橙智慧科技发展有限公司 | Cross-camera pedestrian automatic tracking method |
| CN114842028A (en) * | 2022-05-07 | 2022-08-02 | 深圳先进技术研究院 | Cross-video target tracking method, system, electronic equipment and storage medium |
| CN117237418A (en) * | 2023-11-15 | 2023-12-15 | 成都航空职业技术学院 | Moving object detection method and system based on deep learning |
| CN117237418B (en) * | 2023-11-15 | 2024-01-23 | 成都航空职业技术学院 | Moving object detection method and system based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111462200B (en) | 2023-09-19 |
| WO2021196294A1 (en) | 2021-10-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111462200B (en) | Cross-video pedestrian positioning and tracking method, system and equipment | |
| US11270148B2 (en) | Visual SLAM method and apparatus based on point and line features | |
| JP6095018B2 (en) | Detection and tracking of moving objects | |
| CN107240124B (en) | Cross-lens multi-target tracking method and device based on space-time constraint | |
| CN109829398B (en) | Target detection method in video based on three-dimensional convolution network | |
| US9286678B2 (en) | Camera calibration using feature identification | |
| US9275472B2 (en) | Real-time player detection from a single calibrated camera | |
| CN113313763B (en) | Monocular camera pose optimization method and device based on neural network | |
| JP7147753B2 (en) | Information processing device, information processing method, and program | |
| CN112396656A (en) | Outdoor mobile robot pose estimation method based on fusion of vision and laser radar | |
| Cvišić et al. | Recalibrating the KITTI dataset camera setup for improved odometry accuracy | |
| CN111027462A (en) | Pedestrian track identification method across multiple cameras | |
| Momeni-k et al. | Height estimation from a single camera view | |
| CN112598709B (en) | Pedestrian movement speed intelligent sensing method based on video stream | |
| CN106504274A (en) | A kind of visual tracking method and system based under infrared camera | |
| Liao et al. | SE-Calib: Semantic Edge-Based LiDAR–Camera Boresight Online Calibration in Urban Scenes | |
| CN111829522B (en) | Instant positioning and map construction method, computer equipment and device | |
| CN117593650B (en) | Moving point filtering vision SLAM method based on 4D millimeter wave radar and SAM image segmentation | |
| Xiao et al. | Geo-spatial aerial video processing for scene understanding and object tracking | |
| CN117036404A (en) | Monocular thermal imaging simultaneous positioning and mapping method and system | |
| Sen et al. | SceneCalib: Automatic targetless calibration of cameras and LiDARs in autonomous driving | |
| Revaud et al. | Robust automatic monocular vehicle speed estimation for traffic surveillance | |
| CN116259001A (en) | Multi-view fusion three-dimensional pedestrian posture estimation and tracking method | |
| CN113850864B (en) | GNSS/LIDAR loop detection method for outdoor mobile robot | |
| CN116128919A (en) | Multi-temporal image abnormal target detection method and system based on polar constraint |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |