CN117278837B - Emergency rescue-oriented imaging equipment control method - Google Patents
Emergency rescue-oriented imaging equipment control method Download PDFInfo
- Publication number
- CN117278837B CN117278837B CN202311527002.1A CN202311527002A CN117278837B CN 117278837 B CN117278837 B CN 117278837B CN 202311527002 A CN202311527002 A CN 202311527002A CN 117278837 B CN117278837 B CN 117278837B
- Authority
- CN
- China
- Prior art keywords
- dumping
- target
- image frame
- vector
- pedestrian
- 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 31
- 238000003384 imaging method Methods 0.000 title claims abstract description 11
- 239000013598 vector Substances 0.000 claims abstract description 121
- 230000008859 change Effects 0.000 claims abstract description 46
- 208000027418 Wounds and injury Diseases 0.000 claims abstract description 12
- 230000006378 damage Effects 0.000 claims abstract description 12
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 208000014674 injury Diseases 0.000 claims abstract description 12
- 238000012216 screening Methods 0.000 claims abstract description 7
- 238000013507 mapping Methods 0.000 claims description 12
- 230000001276 controlling effect Effects 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 4
- 230000001105 regulatory effect Effects 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/61—Control of cameras or camera modules based on recognised objects
- H04N23/611—Control of cameras or camera modules based on recognised objects where the recognised objects include parts of the human body
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/48—Extraction of image or video features by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/67—Focus control based on electronic image sensor signals
- H04N23/671—Focus control based on electronic image sensor signals in combination with active ranging signals, e.g. using light or sound signals emitted toward objects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/695—Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The application relates to the technical field of equipment control, in particular to an emergency rescue oriented imaging equipment control method, which comprises the following steps: acquiring human body bounding boxes and head bounding boxes of pedestrians in the current gun camera image frame according to the target detection network; screening pedestrians according to the number of stair steps covered by the human body surrounding frame to obtain target pedestrians; acquiring dumping vectors of a target pedestrian in all gun camera image frames in a current gun camera image frame and a set time period later, and calculating a dumping change rate to determine a dumping result of the target pedestrian; and responding to the dumping result to obtain the position information of the target pedestrian, and controlling the angle and the focal length of the dome camera to acquire the dumping image of the target pedestrian. Through the technical scheme of this application, can be after the pedestrian emptys, control camera equipment gathers pedestrian's image information to confirm pedestrian's injury, and then formulate reasonable emergency rescue measure.
Description
Technical Field
The application relates to the technical field of equipment control, in particular to an emergency rescue oriented imaging equipment control method.
Background
With the continuous promotion of the urban process, the number of supermarkets, rail transit and office buildings is increased, and stairs are indispensable equipment in the places. The stairs can greatly improve the flowing efficiency of pedestrians and bring great convenience for the pedestrians to travel; however, in recent years, stair accidents such as pedestrian dumping, pedestrian treading and the like clearly indicate that the stair has potential safety hazards and is an accident high-rise place; the current research is mainly focused on preventing accidents before the accidents happen in the stair area, but neglecting how to perform emergency rescue after the accidents happen in the stair area.
After accidents such as pedestrian dumping and the like occur in a stair area, because the intensity of pedestrians is high, the image information of the dumped pedestrians cannot be obtained, and then the injury of the dumped pedestrians cannot be confirmed, so that reasonable emergency rescue measures cannot be formulated, and inconvenience is brought to emergency rescue.
At present, patent document with the grant publication number CN109283856B discloses a shooting angle adjusting method of a vehicle recorder, which calculates a lens adjusting angle according to a vehicle state signal through a sensor signal, vehicle lamp adjusting angle information and other vehicle state signals, and adjusts a camera of the vehicle recorder according to the calculated lens adjusting angle.
However, although the method can obtain a larger visual field range through adjusting the lens, the image information of the toppling pedestrian cannot be accurately obtained after the pedestrian topples, and further the injury of the toppling pedestrian cannot be confirmed, so that inconvenience is brought to emergency rescue.
Disclosure of Invention
In order to solve the technical problem, the application provides an imaging equipment control method for emergency rescue, after a pedestrian falls, the imaging equipment is controlled to acquire image information of the falling pedestrian, the injury of the falling pedestrian can be confirmed, and reasonable emergency rescue measures are formulated.
The application provides an emergency rescue-oriented imaging equipment control method, wherein the imaging equipment comprises a gun camera and a ball camera, and the method comprises the following steps: acquiring a human body surrounding frame and a head surrounding frame of each pedestrian in a current rifle bolt image frame according to the trained target detection network, wherein the current rifle bolt image frame comprises image information of all pedestrians in a stair area; screening all pedestrians according to the stair steps covered by the human body bounding box to obtain target pedestrians, wherein the target pedestrians are pedestrians with toppling tendency; for a target pedestrian, acquiring a dumping vector of the target pedestrian in the current gun camera image frame, wherein the starting point of the dumping vector is the center point of a human bounding box corresponding to the target pedestrian, and the ending point is the center point of a head bounding box; acquiring dumping vectors of the target pedestrians in all the gun camera image frames in a set time period adjacent to the current gun camera image frame, and calculating the dumping change rate of the target pedestrians; determining a dumping result of the target pedestrian according to the dumping change rate, wherein the dumping result comprises dumping and non-dumping; responding to the dumping result to obtain the position information of the target pedestrian, and controlling the angle and focal length of the dome camera based on the position information to acquire a dumping image of the target pedestrian; the position information is a human body bounding box of the target pedestrian in a gun camera image frame, and the controlling the angle and the focal length of the dome camera based on the position information to acquire the dumping image of the target pedestrian comprises the following steps: the angle of the ball machine is regulated, so that the center point of the human bounding box corresponding to the target pedestrian coincides with the center point of the image frame of the ball machine; acquiring size information of a human bounding box corresponding to the target pedestrian and the image frame of the dome camera, and calculating a target focal length; adjusting the focal length of the dome camera to the target focal length; the target focal length satisfies the relationship:
wherein,and->The minimum allowable value and the maximum allowable value of the focal length of the dome camera are; />And->For focal length->The lower target pedestrian corresponds to the width and height dimensions of the human body surrounding frame, < ->And->For the width and height dimensions of the dome camera image frame, < >>Is the target focal length.
In one embodiment, the step of screening all pedestrians according to the number of stairs covered by the human body surrounding frame to obtain the target pedestrians includes: according to the color segmentation, a stair step area in a gun camera image frame is obtained, hough straight line detection is carried out on the stair step area, and step edges of each stair step are obtained; for a pedestrian, acquiring the number of step edges in a human body surrounding frame corresponding to the pedestrian; and in response to the number of step edges being greater than a set number, taking the pedestrian as a target pedestrian.
In one embodiment, the set number is empirically preset.
In one embodiment, the method for determining the set number includes: constructing a high-order mapping table, wherein the high-order mapping table comprises a plurality of height ranges and a set number corresponding to each height range; determining height information of the target pedestrian according to the center point coordinates and the height information of the human body bounding box corresponding to the target pedestrian; and inquiring the height order mapping table based on the height information to obtain the set number corresponding to the target pedestrian.
In one embodiment, acquiring the dumping vectors of the target pedestrians in all the bolt face image frames within a set time period adjacent to the current bolt face image frame includes: for a target pedestrian, calculating the intersection ratio of a human body bounding box of the target pedestrian in the current camera image frame and all camera image frames in the next adjacent camera image frame, and taking the human body bounding box corresponding to the maximum intersection ratio as an adjacent human body bounding box, wherein the adjacent human body bounding box is the human body bounding box of the target pedestrian in the next adjacent camera image frame; in the next adjacent gun camera image frame, calculating the intersection ratio of a head bounding box and the adjacent human bounding box, and taking a head bounding box with the intersection ratio being more than or equal to 0 as a target head bounding box, wherein at least one target head bounding box is arranged; taking the central point of the adjacent human bounding box as a starting point and the central point of a target head bounding box as an ending point to obtain a candidate dumping vector in the next adjacent gun camera image frame; calculating the direction correlation between the toppling vector and the candidate toppling vector in the current gun camera image frame; after deleting the candidate dumping vector with the direction correlation larger than or equal to a correlation threshold, calculating the modulus change rate between the dumping vector and the candidate dumping vector in the current gun camera image frame; taking the candidate dumping vector corresponding to the minimum module length change rate as the dumping vector of the target pedestrian in the next adjacent gun camera image frame; iteratively acquiring dumping vectors of the target pedestrians in the next adjacent gun camera image frame until the dumping vectors of the target pedestrians in all gun camera image frames in the set time period are acquired; the directional dependence satisfies the relation:
wherein,for the tilting vector in the current bolt face image frame, < >>For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->A directional correlation between; the modulus change rate satisfies the relation:
wherein,for the tilting vector in the current bolt face image frame, < >>For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->The rate of change of the mode length between.
In one embodiment, the method for determining the set time period includes: the starting time of the set time period is the current rifle bolt image frame; after the current bolt face image frame, calculating the module length variation of the dumping vector between any bolt face image frame and the previous adjacent bolt face image frame; and responding to the mode length variation less than 0, and taking the corresponding last adjacent rifle bolt image frame as the ending moment of the set time period.
In one embodiment, the rate of change of pouring satisfies the relationship:
wherein,to set the +.>Tilting vector in each bolt face image frame, < >>To be within a set time periodTilting vector in each bolt face image frame, < >>For the number of frames of the bolt face in a set period of time,/-for the number of frames of the bolt face>Is the rate of change of pouring.
In one embodiment, determining the dumping result of the target pedestrian as a function of the dumping rate of change includes: comparing the rate of change of pouring with a pouring threshold; responsive to the rate of change of dumping being greater than the dumping threshold, the result of dumping of the target pedestrian is dumping; and responsive to the rate of change of dumping being no greater than the dumping threshold, the result of dumping of the target pedestrian is non-dumping.
In one embodiment, after capturing the dump image of the target pedestrian, the method further comprises: formulating an emergency rescue plan based on the dumping image; wherein, the establishing an emergency rescue scheme based on the dumping image comprises: inputting the dumping image into a feature extraction network, and outputting an image vector corresponding to the dumping image, wherein the image vector is used for reflecting the injury of the target pedestrian; calculating the similarity between the image vector and each standard vector, and taking the standard vector corresponding to the maximum value of the similarity as a matching vector; inquiring a rescue scheme table, and taking a rescue scheme corresponding to the matching vector as a rescue scheme of the target pedestrian; the rescue scheme table comprises all standard vectors and rescue schemes corresponding to the standard vectors.
The technical scheme of the application has the following beneficial technical effects:
according to the technical scheme provided by the application, the human body bounding box and the head bounding box of all pedestrians in the gun camera image frame are firstly obtained, vectors, pointing to the center point of the head bounding box, of the center point of the human body bounding box are used as dumping vectors, and whether the pedestrians are dumped or not is accurately judged according to the dumping change rate of the dumping vectors of the pedestrians in a set time period; after the pedestrian falls, the angle and the focal length of the dome camera are controlled to enable the dome camera image frame to contain complete image information of the falling pedestrian, and the area occupied ratio of the falling pedestrian in the dome camera image frame is maximum, so that a falling image of the falling pedestrian is obtained, the injury of the pedestrian can be confirmed, and reasonable emergency rescue measures are formulated.
Further, in the process of accurately judging whether the pedestrian falls according to the falling change rate of the falling vector of the pedestrian in the set time period, the falling vector of the target pedestrian in the next adjacent rifle bolt image frame is accurately obtained according to the mode length change rate and the direction correlation of the falling vector of the target pedestrian in the current rifle bolt image frame and a plurality of candidate falling vectors in the next adjacent rifle bolt image frame, so that the accuracy of the falling change rate is ensured; the dumping change rate reflects the average value of the dumping vectors of the target pedestrians in the set time period, whether the target pedestrians are dumped or not is judged according to the dumping change rate, false detection of single-frame gun camera image frames is avoided, and accuracy of dumping judgment of the target pedestrians is improved.
Drawings
The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. Several embodiments of the present application are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar or corresponding parts and in which:
fig. 1 is a flowchart of an image capturing apparatus control method for emergency rescue according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a higher order mapping table according to an embodiment of the present application;
fig. 3 is a schematic diagram of a dump vector according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one skilled in the art without making any inventive effort, are intended to be within the scope of the present application based on the embodiments herein.
It should be understood that when the terms "first," "second," and the like are used in the claims, specification, and drawings of this application, they are used merely for distinguishing between different objects and not for describing a particular sequential order. The terms "comprises" and "comprising," when used in the specification and claims of this application, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
According to a first aspect of the present application, the present application provides a method for controlling an image capturing apparatus for emergency rescue; the camera shooting equipment comprises a gun camera and a ball camera, wherein the gun camera is a wide-angle camera, and can acquire global images of all pedestrians in the whole stair area; the ball machine is arranged on the cradle head, and the local image of a specific area in the stair area can be obtained by adjusting the angle and the focal length of the ball machine.
Fig. 1 is a flowchart of an image capturing apparatus control method for emergency rescue according to an embodiment of the present application. As shown in fig. 1, the test method 100 includes steps S101 to S106, which are described in detail below.
S101, acquiring a human body bounding box and a head bounding box of each pedestrian in a current rifle bolt image frame according to the trained target detection network, wherein the current rifle bolt image frame comprises image information of all pedestrians in a stair area.
In one embodiment, in order to enable timely establishment of reasonable emergency rescue measures after a pedestrian tip occurs in a stair area, a rifle bolt is deployed at a overlooking view just above the stair area, and the rifle bolt can acquire information of all pedestrians in the whole stair area, that is, the current rifle bolt image frame comprises image information of all pedestrians in the stair area at the current moment.
The input of the target detection network is any image to be detected, and the input is a human body bounding box and a head bounding box of each pedestrian in the image to be detected, wherein the human body bounding box of one pedestrian covers all areas of the pedestrian, and the head bounding box of one pedestrian covers the head area of the pedestrian.
It will be appreciated that the human bounding box and the head bounding box are both represented by the center point coordinates and the width and height dimensions of the bounding box. The object detection network may be any one of the existing object detection networks such as YOLO, SSD, or centrnet.
S102, screening all pedestrians according to the number of stair steps covered by the human body surrounding frame to obtain target pedestrians, wherein the target pedestrians are pedestrians with toppling trends.
In one embodiment, since the rifle bolt is in a top view, when a pedestrian normally passes through, the size of the human body surrounding frame of the pedestrian in the passing direction is smaller, and when the pedestrian has a tilting tendency, the size of the human body surrounding frame corresponding to the pedestrian extends in the passing direction, so that the size of the human body surrounding frame in the passing direction is larger, wherein the passing direction is along the moving direction of the pedestrian on the stairs. The size of the stair steps is fixed, so that the stair steps are used as reference distances, whether the pedestrians have a toppling trend can be judged according to the number of the stair steps covered by the human body surrounding frame, and the pedestrians with the toppling trend are used as target pedestrians.
Specifically, the step number of the stairs covered by the human body surrounding frame is used for screening all pedestrians to obtain target pedestrians, and the step of obtaining the target pedestrians comprises the following steps: according to the color segmentation, a stair step area in a gun camera image frame is obtained, hough straight line detection is carried out on the stair step area, and step edges of each stair step are obtained; for a pedestrian, acquiring the number of step edges in a human body surrounding frame corresponding to the pedestrian; and in response to the number of step edges being greater than a set number, taking the pedestrian as a target pedestrian. Wherein the number of target pedestrians may be 0, one or more.
According to the color difference between the stair steps and the rest areas, the stair areas in the gun camera image frame are obtained, and then the stair edges of each stair step in the stair areas are determined.
It can be understood that when the number of the step edges is greater than the set number, the size of the human body surrounding frame in the passing direction is indicated to be larger, and the inclination trend exists for the pedestrians.
In an alternative embodiment, the set number may be preset empirically, specifically, the set number has a value of 4.
In another alternative embodiment, the pedestrians are different in height, and the set number is related to the height of the target pedestrian in order to accurately acquire the target pedestrian. Specifically, the method for determining the set number includes: constructing a high-order mapping table, wherein the high-order mapping table comprises a plurality of height ranges and a set number corresponding to each height range; determining height information of the target pedestrian according to the center point coordinates and the height information of the human body bounding box corresponding to the target pedestrian; and inquiring the height order mapping table based on the height information to obtain the set number corresponding to the target pedestrian.
Fig. 2 is a schematic diagram of a higher-order mapping table according to an embodiment of the present application; the set number corresponding to any height range can be obtained from the height order mapping table.
The center point coordinates of the human body surrounding frame corresponding to the target pedestrian can reflect the relative position information of the target pedestrian to the gun camera, the height information of the human body surrounding frame can reflect the height of the target pedestrian under the view angle of the relative position information, and the height information of the target pedestrian can be calculated according to the height information of the human body surrounding frame and the center point coordinates.
S103, for a target pedestrian, acquiring a dumping vector of the target pedestrian in the current gun camera image frame, wherein the starting point of the dumping vector is the center point of a human bounding box corresponding to the target pedestrian, and the ending point is the center point of a head bounding box.
In one embodiment, the number of target pedestrians is 0, 1 or more in the current bolt face image frame. FIG. 3 is a schematic diagram of a dump vector according to an embodiment of the present application; for any pedestrian, the center points of the human body bounding box and the head bounding box corresponding to the pedestrian are obtained, the center point of the head bounding box is taken as a starting point, the center point of the human body bounding box is taken as an ending point, and the dumping vector of the pedestrian in the current gun camera image frame is constructed. It will be appreciated that the greater the modulus of the dump vector, the greater the likelihood of dumping for the corresponding pedestrian.
S104, obtaining dumping vectors of the target pedestrians in all the gun camera image frames in a set time period adjacent to the current gun camera image frame, and calculating the dumping change rate of the target pedestrians.
In one embodiment, the toppling vector of the target pedestrian in the current bolt face image frame is in an instantaneous state and is greatly influenced by the gesture information of the target pedestrian, and it is difficult to accurately judge whether the target pedestrian topples or not only by means of the toppling vector of the target pedestrian in the current bolt face image frame, so that the information of the bolt face image frame in a set time period adjacent to the current bolt face image frame needs to be combined to accurately judge whether the target pedestrian topples or not.
Specifically, acquiring the dumping vectors of the target pedestrians in all the camera image frames in a set time period adjacent to the current camera image frame comprises: for a target pedestrian, calculating the intersection ratio of a human body bounding box of the target pedestrian in the current camera image frame and all camera image frames in the next adjacent camera image frame, and taking the human body bounding box corresponding to the maximum intersection ratio as an adjacent human body bounding box, wherein the adjacent human body bounding box is the human body bounding box of the target pedestrian in the next adjacent camera image frame; in the next adjacent gun camera image frame, calculating the intersection ratio of a head bounding box and the adjacent human bounding box, and taking a head bounding box with the intersection ratio being more than or equal to 0 as a target head bounding box, wherein at least one target head bounding box is arranged; taking the central point of the adjacent human bounding box as a starting point and the central point of a target head bounding box as an ending point to obtain a candidate dumping vector in the next adjacent gun camera image frame; calculating the direction correlation between the toppling vector and the candidate toppling vector in the current gun camera image frame; after deleting the candidate dumping vector with the direction correlation larger than or equal to a correlation threshold, calculating the modulus change rate between the dumping vector and the candidate dumping vector in the current gun camera image frame; taking the candidate dumping vector corresponding to the minimum module length change rate as the dumping vector of the target pedestrian in the next adjacent gun camera image frame; and iteratively acquiring the dumping vectors of the target pedestrians in the next adjacent gun camera image frame until the dumping vectors of the target pedestrians in all gun camera image frames in the set time period are acquired. Wherein, the correlation threshold value is 0.1.
It will be appreciated that, because the stair area is a dense area of people, in a dense situation, the human bounding box of the target pedestrian in the next adjacent bolt face image frame may intersect with the head bounding boxes, i.e. the number of the target head bounding boxes is 1 or more, so as to accurately obtain the dumping vector of the target pedestrian in the next adjacent bolt face image frame, and the plurality of candidate dumping vectors are screened according to the mode length change rate and the direction correlation of the dumping vector of the target pedestrian in the current bolt face image frame and the candidate dumping vector in the next adjacent bolt face image frame.
Wherein the directional correlation satisfies the relation:
wherein,for the tilting vector in the current bolt face image frame, < >>For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->Directional correlation between them.
Wherein the modulus change rate satisfies the relationship:
wherein,for the tilting vector in the current bolt face image frame, < >>For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->The rate of change of the mode length between.
In one embodiment, the method for determining the set time period includes: the starting time of the set time period is the current rifle bolt image frame; after the current bolt face image frame, calculating the module length variation of the dumping vector between any bolt face image frame and the previous adjacent bolt face image frame; and responding to the mode length variation less than 0, and taking the corresponding last adjacent rifle bolt image frame as the ending moment of the set time period.
In one embodiment, after obtaining the dumping vectors of the target pedestrians in all the camera image frames in the set time period adjacent to the current camera image frame, the dumping change rate of the target pedestrians can be calculated, and the dumping change rate satisfies the relation:
wherein,to set the +.>Tilting vector in each bolt face image frame, < >>To be within a set time periodTilting vector in each bolt face image frame, < >>For the number of frames of the bolt face in a set period of time,/-for the number of frames of the bolt face>Is the rate of change of pouring.
It can be understood that the dumping change rate reflects the average value of the dumping vectors of the target pedestrians in the set time period, so that false detection of single-frame gun camera image frames is avoided, and the accuracy of dumping judgment of the target pedestrians is improved.
S105, determining the dumping results of the target pedestrians according to the dumping change rate, wherein the dumping results comprise dumping and non-dumping.
In one embodiment, the rate of change of pouring is compared to a pouring threshold; responsive to the rate of change of dumping being greater than the dumping threshold, the result of dumping of the target pedestrian is dumping; and responsive to the rate of change of dumping being no greater than the dumping threshold, the result of dumping of the target pedestrian is non-dumping. Wherein the dumping threshold is 0.6.
And S106, responding to the dumping result to be dumping, acquiring the position information of the target pedestrian, and controlling the angle and the focal length of the ball machine based on the position information so as to acquire a dumping image of the target pedestrian.
In one embodiment, in response to the dumping result being dumping, indicating that a target pedestrian in the stair area has a dumping accident, in order to accurately determine the injury of the target pedestrian after dumping, the angle and the focal length of the dome camera need to be controlled to acquire a dumping image of the target pedestrian.
Specifically, the position information is a human body bounding box of the target pedestrian in a gun camera image frame, and the controlling the angle and the focal length of the dome camera based on the position information to acquire the dumping image of the target pedestrian comprises: the angle of the ball machine is regulated, so that the center point of the human bounding box corresponding to the target pedestrian coincides with the center point of the image frame of the ball machine; acquiring size information of a human bounding box corresponding to the target pedestrian and the image frame of the dome camera, and calculating a target focal length; and adjusting the focal length of the dome camera to the target focal length.
The larger the focal length is, the larger the human body surrounding frame of the target pedestrian in the image frame of the dome camera is, so that when the human body surrounding frame of the target pedestrian occupies most of the area of the image frame of the dome camera, a toppling image of the target pedestrian can be obtained, and the toppling image can accurately and clearly reflect the injury situation of the target pedestrian after toppling.
Specifically, the target focal length satisfies the relation:
wherein,and->The minimum allowable value and the maximum allowable value of the focal length of the dome camera are; />And->For focal length->The lower target pedestrian corresponds to the width and height dimensions of the human body surrounding frame, < ->And->For the width and height dimensions of the dome camera image frame, < >>Is the target focal length.
It will be appreciated that the number of components,and after the focal length is regulated by constraint conditions, the image frame of the spherical camera contains the image information of the complete target pedestrian, and the area occupied by the target pedestrian in the image frame of the spherical camera is the largest.
Therefore, the control of the camera equipment under the emergency rescue scene is realized, the dumping image of the dumping pedestrian can be acquired clearly and accurately, the dumping image can accurately reflect the injury situation of the dumping pedestrian, and further reasonable emergency rescue measures can be formulated.
In an alternative embodiment, after capturing the dump image of the target pedestrian, the method further comprises: formulating an emergency rescue plan based on the dumping image; wherein, the establishing an emergency rescue scheme based on the dumping image comprises: inputting the dumping image into a feature extraction network, and outputting an image vector corresponding to the dumping image, wherein the image vector is used for reflecting the injury of the target pedestrian; calculating the similarity between the image vector and each standard vector, and taking the standard vector corresponding to the maximum value of the similarity as a matching vector; inquiring a rescue scheme table, and taking a rescue scheme corresponding to the matching vector as a rescue scheme of the target pedestrian; the rescue scheme table comprises all standard vectors and rescue schemes corresponding to the standard vectors.
The rescue scheme table is preset, the feature extraction network is a self-coding network, and the image vector corresponding to the dumping image is an output result of an encoder in the self-coding network.
Technical principles and implementation details of the emergency rescue oriented imaging device control method of the present application are described above through specific embodiments. According to the technical scheme provided by the application, the human body bounding box and the head bounding box of all pedestrians in the gun camera image frame are firstly obtained, vectors, pointing to the center point of the head bounding box, of the center point of the human body bounding box are used as dumping vectors, and whether the pedestrians are dumped or not is accurately judged according to the dumping change rate of the dumping vectors of the pedestrians in a set time period; after the pedestrian falls, the angle and the focal length of the dome camera are controlled to enable the dome camera image frame to contain complete image information of the falling pedestrian, and the area occupied ratio of the falling pedestrian in the dome camera image frame is maximum, so that a falling image of the falling pedestrian is obtained, the injury of the pedestrian can be confirmed, and reasonable emergency rescue measures are formulated.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the claims. It should be noted that variations and modifications can be made by those skilled in the art without departing from the spirit of the present application, which falls within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.
Claims (7)
1. An emergency rescue oriented imaging apparatus control method, the imaging apparatus including a bolt and a ball machine, comprising:
acquiring a human body surrounding frame and a head surrounding frame of each pedestrian in a current rifle bolt image frame according to the trained target detection network, wherein the current rifle bolt image frame comprises image information of all pedestrians in a stair area;
screening all pedestrians according to the stair steps covered by the human body bounding box to obtain target pedestrians, wherein the target pedestrians are pedestrians with toppling tendency;
for a target pedestrian, acquiring a dumping vector of the target pedestrian in the current gun camera image frame, wherein the starting point of the dumping vector is the center point of a human bounding box corresponding to the target pedestrian, and the ending point is the center point of a head bounding box;
acquiring dumping vectors of the target pedestrians in all the gun camera image frames in a set time period adjacent to the current gun camera image frame, and calculating the dumping change rate of the target pedestrians;
determining a dumping result of the target pedestrian according to the dumping change rate, wherein the dumping result comprises dumping and non-dumping;
responding to the dumping result to obtain the position information of the target pedestrian, and controlling the angle and focal length of the dome camera based on the position information to acquire a dumping image of the target pedestrian;
the position information is a human body bounding box of the target pedestrian in a gun camera image frame, and the controlling the angle and the focal length of the dome camera based on the position information to acquire the dumping image of the target pedestrian comprises the following steps: the angle of the ball machine is regulated, so that the center point of the human bounding box corresponding to the target pedestrian coincides with the center point of the image frame of the ball machine; acquiring size information of a human bounding box corresponding to the target pedestrian and the image frame of the dome camera, and calculating a target focal length; adjusting the focal length of the dome camera to the target focal length; the target focal length satisfies the relationship:
wherein,and->The minimum allowable value and the maximum allowable value of the focal length of the dome camera are; />And->For focal length->The lower target pedestrian corresponds to the width and height dimensions of the human body surrounding frame, < ->And->For the width and height dimensions of the dome camera image frame, < >>Is the target focal length;
the acquiring the dumping vectors of the target pedestrians in all the gun camera image frames in the adjacent set time period after the current gun camera image frame comprises the following steps:
for a target pedestrian, calculating the intersection ratio of a human body bounding box of the target pedestrian in the current camera image frame and all camera image frames in the next adjacent camera image frame, and taking the human body bounding box corresponding to the maximum intersection ratio as an adjacent human body bounding box, wherein the adjacent human body bounding box is the human body bounding box of the target pedestrian in the next adjacent camera image frame;
in the next adjacent gun camera image frame, calculating the intersection ratio of a head bounding box and the adjacent human bounding box, and taking a head bounding box with the intersection ratio being more than or equal to 0 as a target head bounding box, wherein at least one target head bounding box is arranged;
taking the central point of the adjacent human bounding box as a starting point and the central point of a target head bounding box as an ending point to obtain a candidate dumping vector in the next adjacent gun camera image frame;
calculating the direction correlation between the toppling vector and the candidate toppling vector in the current gun camera image frame;
after deleting the candidate dumping vector with the direction correlation larger than or equal to a correlation threshold, calculating the modulus change rate between the dumping vector and the candidate dumping vector in the current gun camera image frame;
taking the candidate dumping vector corresponding to the minimum module length change rate as the dumping vector of the target pedestrian in the next adjacent gun camera image frame;
iteratively acquiring dumping vectors of the target pedestrians in the next adjacent gun camera image frame until the dumping vectors of the target pedestrians in all gun camera image frames in the set time period are acquired;
the directional dependence satisfies the relation:
wherein,for the current bolt face image frameToppling vector in->For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->A directional correlation between;
the modulus change rate satisfies the relation:
wherein,for the tilting vector in the current bolt face image frame, < >>For a candidate dump vector in the next adjacent bolt face image frame,/for each candidate dump vector>Is->And->A rate of change of the mold length therebetween;
the pouring change rate satisfies the relation:
wherein,to set the +.>Tilting vector in each bolt face image frame, < >>To set the +.>Tilting vector in each bolt face image frame, < >>For the number of frames of the bolt face in a set period of time,/-for the number of frames of the bolt face>Is the rate of change of pouring.
2. The method for controlling an image capturing apparatus for emergency rescue according to claim 1, wherein the step of screening all pedestrians according to the number of stairs covered by the human body bounding box to obtain a target pedestrian comprises:
according to the color segmentation, a stair step area in a gun camera image frame is obtained, hough straight line detection is carried out on the stair step area, and step edges of each stair step are obtained;
for a pedestrian, acquiring the number of step edges in a human body surrounding frame corresponding to the pedestrian;
and in response to the number of step edges being greater than a set number, taking the pedestrian as a target pedestrian.
3. The image capturing apparatus control method for emergency rescue according to claim 2, wherein the set number is set in advance based on experience.
4. The image capturing apparatus control method for emergency rescue according to claim 2, wherein the determination method for the set number includes:
constructing a high-order mapping table, wherein the high-order mapping table comprises a plurality of height ranges and a set number corresponding to each height range;
determining height information of the target pedestrian according to the center point coordinates and the height information of the human body bounding box corresponding to the target pedestrian;
and inquiring the height order mapping table based on the height information to obtain the set number corresponding to the target pedestrian.
5. The image capturing apparatus control method for emergency rescue according to claim 1, wherein the determination method for the set period of time includes:
the starting time of the set time period is the current rifle bolt image frame; after the current bolt face image frame, calculating the module length variation of the dumping vector between any bolt face image frame and the previous adjacent bolt face image frame;
and responding to the mode length variation less than 0, and taking the corresponding last adjacent rifle bolt image frame as the ending moment of the set time period.
6. The image capturing apparatus control method for emergency rescue according to claim 1, wherein determining the dumping result of the target pedestrian in accordance with the dumping change rate includes:
comparing the rate of change of pouring with a pouring threshold;
responsive to the rate of change of dumping being greater than the dumping threshold, the result of dumping of the target pedestrian is dumping;
and responsive to the rate of change of dumping being no greater than the dumping threshold, the result of dumping of the target pedestrian is non-dumping.
7. The image capturing apparatus control method for emergency rescue according to any one of claims 1 to 6, wherein after collecting a dump image of a target pedestrian, the method further comprises:
formulating an emergency rescue plan based on the dumping image;
wherein, the establishing an emergency rescue scheme based on the dumping image comprises: inputting the dumping image into a feature extraction network, and outputting an image vector corresponding to the dumping image, wherein the image vector is used for reflecting the injury of the target pedestrian; calculating the similarity between the image vector and each standard vector, and taking the standard vector corresponding to the maximum value of the similarity as a matching vector; inquiring a rescue scheme table, and taking a rescue scheme corresponding to the matching vector as a rescue scheme of the target pedestrian;
the rescue scheme table comprises all standard vectors and rescue schemes corresponding to the standard vectors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311527002.1A CN117278837B (en) | 2023-11-16 | 2023-11-16 | Emergency rescue-oriented imaging equipment control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311527002.1A CN117278837B (en) | 2023-11-16 | 2023-11-16 | Emergency rescue-oriented imaging equipment control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117278837A CN117278837A (en) | 2023-12-22 |
| CN117278837B true CN117278837B (en) | 2024-01-26 |
Family
ID=89204753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311527002.1A Active CN117278837B (en) | 2023-11-16 | 2023-11-16 | Emergency rescue-oriented imaging equipment control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117278837B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105979210A (en) * | 2016-06-06 | 2016-09-28 | 深圳市深网视界科技有限公司 | Pedestrian identification system based on multi-ball multi-gun camera array |
| KR101773898B1 (en) * | 2017-02-16 | 2017-09-01 | (주)하이디어 솔루션즈 | System for monitoring fall of care-object |
| CN109376577A (en) * | 2018-08-27 | 2019-02-22 | 中国电子科技集团公司电子科学研究院 | Video synthesis analysis method and system |
| CN109584309A (en) * | 2018-11-16 | 2019-04-05 | 厦门博聪信息技术有限公司 | A kind of twin-lens emergency cloth ball-handling of rifle ball linkage |
| KR20190060249A (en) * | 2017-11-24 | 2019-06-03 | 사단법인 캠틱종합기술원 | Method for dropping rescue equipment and drone for rescue using the same |
| CN109919132A (en) * | 2019-03-22 | 2019-06-21 | 广东省智能制造研究所 | A kind of pedestrian's tumble recognition methods based on skeleton detection |
| CN111724566A (en) * | 2020-05-20 | 2020-09-29 | 同济大学 | Pedestrian falling detection method and device based on intelligent lamp pole video monitoring system |
| JP2021117904A (en) * | 2020-01-29 | 2021-08-10 | 積水ハウス株式会社 | Fall detection device |
| WO2022041484A1 (en) * | 2020-08-26 | 2022-03-03 | 歌尔股份有限公司 | Human body fall detection method, apparatus and device, and storage medium |
| KR102389285B1 (en) * | 2021-07-12 | 2022-04-22 | 주식회사 씨앤에이아이 | Method for detecting a falling accident based on machine learning and system using the same |
| CN114565974A (en) * | 2022-02-28 | 2022-05-31 | 安徽兰臣信息科技有限公司 | Human body dumping identification algorithm based on 3D vision |
| CN115116127A (en) * | 2022-05-25 | 2022-09-27 | 西安北斗安全技术有限公司 | Fall detection method based on computer vision and artificial intelligence |
| CN115311591A (en) * | 2021-12-09 | 2022-11-08 | 北京市基础设施投资有限公司 | Early warning method and device for abnormal behaviors and intelligent camera |
| CN116052004A (en) * | 2023-02-17 | 2023-05-02 | 深圳金三立视频科技股份有限公司 | Bidirectional monitoring method and device for abnormal events, electronic equipment and storage medium |
| CN116246343A (en) * | 2023-02-16 | 2023-06-09 | 深圳英飞拓智能技术有限公司 | Light human body behavior recognition method and device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9152856B2 (en) * | 2013-12-19 | 2015-10-06 | Institute For Information Industry | Pedestrian detection system and method |
| US20190365290A1 (en) * | 2018-05-31 | 2019-12-05 | Medtronic, Inc. | Body and body part orientation and posture monitoring |
| US20230004740A1 (en) * | 2021-06-30 | 2023-01-05 | Ubtech North America Research And Development Center Corp | Human lying posture detection method and mobile machine using the same |
-
2023
- 2023-11-16 CN CN202311527002.1A patent/CN117278837B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105979210A (en) * | 2016-06-06 | 2016-09-28 | 深圳市深网视界科技有限公司 | Pedestrian identification system based on multi-ball multi-gun camera array |
| KR101773898B1 (en) * | 2017-02-16 | 2017-09-01 | (주)하이디어 솔루션즈 | System for monitoring fall of care-object |
| KR20190060249A (en) * | 2017-11-24 | 2019-06-03 | 사단법인 캠틱종합기술원 | Method for dropping rescue equipment and drone for rescue using the same |
| CN109376577A (en) * | 2018-08-27 | 2019-02-22 | 中国电子科技集团公司电子科学研究院 | Video synthesis analysis method and system |
| CN109584309A (en) * | 2018-11-16 | 2019-04-05 | 厦门博聪信息技术有限公司 | A kind of twin-lens emergency cloth ball-handling of rifle ball linkage |
| CN109919132A (en) * | 2019-03-22 | 2019-06-21 | 广东省智能制造研究所 | A kind of pedestrian's tumble recognition methods based on skeleton detection |
| JP2021117904A (en) * | 2020-01-29 | 2021-08-10 | 積水ハウス株式会社 | Fall detection device |
| CN111724566A (en) * | 2020-05-20 | 2020-09-29 | 同济大学 | Pedestrian falling detection method and device based on intelligent lamp pole video monitoring system |
| WO2022041484A1 (en) * | 2020-08-26 | 2022-03-03 | 歌尔股份有限公司 | Human body fall detection method, apparatus and device, and storage medium |
| KR102389285B1 (en) * | 2021-07-12 | 2022-04-22 | 주식회사 씨앤에이아이 | Method for detecting a falling accident based on machine learning and system using the same |
| CN115311591A (en) * | 2021-12-09 | 2022-11-08 | 北京市基础设施投资有限公司 | Early warning method and device for abnormal behaviors and intelligent camera |
| CN114565974A (en) * | 2022-02-28 | 2022-05-31 | 安徽兰臣信息科技有限公司 | Human body dumping identification algorithm based on 3D vision |
| CN115116127A (en) * | 2022-05-25 | 2022-09-27 | 西安北斗安全技术有限公司 | Fall detection method based on computer vision and artificial intelligence |
| CN116246343A (en) * | 2023-02-16 | 2023-06-09 | 深圳英飞拓智能技术有限公司 | Light human body behavior recognition method and device |
| CN116052004A (en) * | 2023-02-17 | 2023-05-02 | 深圳金三立视频科技股份有限公司 | Bidirectional monitoring method and device for abnormal events, electronic equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| Bo-Hua Wang,et al..Fall Detection Based on Dual-Channel Feature Integration.《 IEEE Access ( Volume: 8)》.2020,全文. * |
| Thang Nguyen Dang,et al..Fast and Accurate Fall Detection and Warning System Using Image Processing Technology.《2021 International Conference on Advanced Technologies for Communications (ATC) 》.2021,全文. * |
| 田一明..基于加速度时域特征和Adaboost-SVM级联分类器的跌倒检测研究.《中南民族大学学报(自然科学版)》.2018,全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117278837A (en) | 2023-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110032962B (en) | Object detection method, device, network equipment and storage medium | |
| JP3503230B2 (en) | Nighttime vehicle recognition device | |
| KR102105162B1 (en) | A smart overspeeding vehicle oversee apparatus for analyzing vehicle speed, vehicle location and traffic volume using radar, for detecting vehicles that violate the rules, and for storing information on them as videos and images, a smart traffic signal violation vehicle oversee apparatus for the same, and a smart city solution apparatus for the same | |
| KR100578041B1 (en) | Network camera apparatus having a traffic information collection and addition function and the traffic information process method | |
| KR101845943B1 (en) | A system and method for recognizing number plates on multi-lane using one camera | |
| JP2003168104A (en) | Recognition device of white line on road | |
| CN111832536A (en) | Lane line detection method and device | |
| JP2001216519A (en) | Traffic monitor device | |
| JP3729025B2 (en) | Pedestrian detection device | |
| JP2007078358A (en) | Measurement device for mobile body, signal control system, measurement method of mobile body, and computer program | |
| Gulati et al. | Image processing in intelligent traffic management | |
| CN111856611A (en) | Trigger detection method, device and system | |
| JP2018055597A (en) | Vehicle type discrimination device and vehicle type discrimination method | |
| KR101112562B1 (en) | Image detecting apparatus and the method thereof | |
| KR102100965B1 (en) | System for analyzing vehicle information on multi-lane embedded a camera and analyzing method using thereof | |
| KR101026778B1 (en) | Vehicle image detection apparatus | |
| CN117278837B (en) | Emergency rescue-oriented imaging equipment control method | |
| KR101073053B1 (en) | Auto Transportation Information Extraction System and Thereby Method | |
| KR102204015B1 (en) | Lifting gate based on image and vehicle control method through the same | |
| CN112241004B (en) | Object recognition device | |
| JP4788399B2 (en) | Pedestrian detection method, apparatus, and program | |
| CN116402994A (en) | Railway danger monitoring method based on laser radar and video image fusion | |
| KR101154350B1 (en) | Multiple Images Processing system using Detection of Objects and segmenting - recognizing of High- resolution Image and there of. | |
| KR102484691B1 (en) | Vehicle detection system and vehicle detection method using stereo camera and radar | |
| KR101026852B1 (en) | Method and system of supervision for car stop or parking |
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 |