CN104298353A - Inverse kinematics based vehicle monitoring and burglary preventing method and system - Google Patents
Inverse kinematics based vehicle monitoring and burglary preventing method and system Download PDFInfo
- Publication number
- CN104298353A CN104298353A CN201410525013.0A CN201410525013A CN104298353A CN 104298353 A CN104298353 A CN 104298353A CN 201410525013 A CN201410525013 A CN 201410525013A CN 104298353 A CN104298353 A CN 104298353A
- Authority
- CN
- China
- Prior art keywords
- human body
- action
- burglary
- skeleton
- data
- 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.)
- Pending
Links
Classifications
-
- 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
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
Abstract
The invention provides an inverse kinematics based vehicle monitoring and burglary preventing method. The method comprises the following steps: 1, catching the action of a human body within a field of view through a camera mounted in a shed on real time; 2, obtaining data stream through an interface provided by the inverse kinematics software, and extracting x, y and z coordinate data of each joint node of a human body skeleton model; 3, calculating direction vectors of the parts of the human body and the characteristic parameters of the parts according to the x, y and z coordinates of each joint node of the human body model, and limiting the parameters to a virtual human body model of a control interface; 4, matching and comparing the virtual human body action characteristics with the suspected burglary action characteristic library; once the similarity reaches the threshold h, initially determining the suspected burglary action; 5, starting the monitoring warming program when the human body action is recognized as the suspected burglary action, and then performing real-time warning. The invention also provides an inverse kinematics based vehicle monitoring and burglary preventing system. With the adoption of the method and system, the action recognizing can be effectively performed, the real-time performance is improved, and the field alarming function is realized.
Description
Technical field
The invention belongs to vehicle monitoring field, particularly relate to a kind of vehicle monitoring theft preventing method and system.
Background technology
In current campus, often there is bicycle, electric motor car larceny case in the bicycle of residential quarters and megastore, electric motor car garage.
Due to the function that common monitoring camera does not possess action recognition and reports to the police then and there, monitoring video can only be transferred to check after robber's car case occurs afterwards, and now often lost the best and solve a case opportunity, also do not have so far a kind ofly finding the method that the behavior of stealing of doubtful car just can be reported to the police or system invention then and there out.
Summary of the invention
Do not possess action recognition and warning function, real-time are poor deficiency then and there in order to what overcome existing vehicle monitoring mode, the invention provides and a kind ofly effectively carry out the vehicle monitoring theft preventing method based on inverse kinematics and the system that action recognition, real-time are good, have warning function then and there.
The technical solution adopted for the present invention to solve the technical problems is:
Based on a vehicle monitoring theft preventing method for inverse kinematics, described method comprises the steps:
Step 1, the Kinect video camera that bicycle shed is installed catches the action of human body in the visual field in real time, and the data of acquisition are transferred to coupled PC;
Step 2, PC obtains Skeleton Frame data stream by the interface that inverse kinematics software Kinect SDK provides and extracts the x, y, z coordinate data of each articulation point of human skeleton model.
Step 3, calculates the characteristic parameter between the direction vector of human body and position, by the virtual human model of restriction on the parameters in control inerface by the x, y, z coordinate of each articulation point of manikin;
Step 4, virtual human body motion characteristic and default " doubtful motion characteristic storehouse stolen by car " are carried out matching ratio comparatively, if when the difference parameter of virtual human body motion characteristic weighting reaches threshold value h with the similarity presetting the car stealing data characteristics storehouse and steal doubtful motion characteristic, be then that doubtful action stolen by car by preliminary judgement;
Step 5, monitors warning program and starts, carry out Real-time Alarm when human action is identified as " doubtful action stolen by car ".
Further, in described step 5, the dio Output Modules first causing control center to be connected with PC gives the alarm the sound, then according to the accuracy of identification data, is cut off report to the police or connect on-the-spot speaker unit and report to the police by operating personnel.
The antitheft burglary-resisting system of vehicle monitoring based on inverse kinematics, described burglary-resisting system comprises Kinect camera, motion capture module, action recognition module and alarm module,
Described motion capture module, receives the data of Kinect camera, and application Kinect somatosensory device, draws the 3 d pose of human body by analyzing skeleton structure position model;
Described action recognition module, utilizes turning axle representation rotation to be expressed as the vectorial N of 1 × 3, rotating vector N represent around turning axle direction a anglec of rotation θ, wherein θ=|| a|| is the mould of N, a=[a
x, a
y, a
z]
tfor normalization sense of rotation, meet N=θ a, corresponding rotation matrix R
n(3 × 3) draw with Rodrigues formulae discovery;
The root node of described human body degree of freedom is positioned at the hip of health, its state modulator human skeleton position in world coordinate system, has 3 and rotates 3 translations, totally 6 degree of freedom; Translation rotation parameter is applied in the manikin set up in advance, draws the motion characteristic of simulation human body, carries out right, if when similarity reaches threshold value h, be tentatively decided to be car and steal doubtful action with the theft data characteristics storehouse of presetting;
Described alarm module, monitors warning program and starts, carry out Real-time Alarm when human action is identified as " doubtful action stolen by car ".
Further, described alarm module comprises dio Output Modules and GPRS alarm module.
Further again, in described motion capture module, obtain the color image data of pedestrian's posture in garage, skeleton data and depth image data by colour imagery shot, infrared camera and image acquisition device in Kinect somatosensory device; Be 20 D coordinates value of joint in system spatial coordinates system in skeleton by the color image data of acquisition, skeleton data and depth image data transformations; Human body skeleton pattern can be expressed as a tree structure about node, represents a joint with each node.
In described action recognition module, human body degree of freedom is the parameter that can independently change in skeleton, in the mutual independently situation in each joint of hypothesis, equals articulate degree of freedom summation; For picture elbow, knee revolute joint, distribute 1 degree of freedom, turning axle is fixed as x-axis; For picture shoulder, hip ball groove joint model, turning axle is positioned at any direction of sphere, adds 1 degree of freedom pivoted, is divided into and joins 3 degree of freedom, for neck joint, then distributes 2 degree of freedom, and restriction turning axle is positioned at z=0 plane.
In described action recognition module, described theft data characteristics storehouse is that scene is carried out constantly perfect by artificial simulation or existing robber garage, and the positive sample of not open close mistake and the continuous Real time identification of negative sample are optimized with identifying.
In described theft data characteristics storehouse, adopt the identification optimization method based on hidden Markov model.
Beneficial effect of the present invention is mainly manifested in: effectively carry out action recognition, real-time good, there is warning function then and there.
Accompanying drawing explanation
Fig. 1 is the basic framework figure of vehicle monitoring burglary-resisting system.
Fig. 2 is the flow chart of steps of vehicle monitoring burglary-resisting system.
Fig. 3 is the process flow diagram of vehicle monitoring theft preventing method action recognition process.
Fig. 4 is the schematic diagram of the tree-shaped skeleton model parameter that inverse kinematics software need record.
Fig. 5 is the schematic diagram of human skeleton model in Activity recognition feature.
Fig. 6 is that reality scene feeds back the simulation human action figure obtained.
In figure: 1, motion capture module, 2 action recognition modules, 3 loudspeakers, 4, GPRS alarm module.
Embodiment
Below in conjunction with accompanying drawing, the invention will be further described.
With reference to Fig. 1 ~ Fig. 6, a kind of vehicle monitoring theft preventing method based on inverse kinematics, comprises the steps:
Step 1, the Kinect video camera that bicycle shed is installed catches the action of human body in the visual field in real time, and the data of acquisition are transferred to coupled PC.
Step 2, PC obtains Skeleton Frame data stream by the interface that inverse kinematics software Kinect SDK provides and extracts the x, y, z coordinate data of human skeleton model (see accompanying drawing 5) each articulation point.
Step 3, calculates the characteristic parameter between the direction vector of human body and position, by the virtual human model of restriction on the parameters in control inerface by the x, y, z coordinate of each articulation point of manikin.
Step 4, virtual human body motion characteristic is carried out as shown in Figure 6 matching ratio with " doubtful motion characteristic storehouse stolen by car " comparatively (adopting the identification optimization method based on hidden Markov model), if when the difference parameter of virtual human body motion characteristic weighting reaches threshold value h with the similarity presetting the car stealing data characteristics storehouse and steal doubtful motion characteristic, be then that doubtful action stolen by car by preliminary judgement.
Step 5, monitors warning program and starts, carry out Real-time Alarm when human action is identified as " doubtful action stolen by car ".
Further, in described step 5, the dio Output Modules 3 first causing control center to be connected with PC gives the alarm the sound, then according to the accuracy of identification data, is cut off report to the police or connect on-the-spot speaker unit and report to the police by operating personnel.
With reference to Fig. 1, a kind of vehicle monitoring burglary-resisting system based on inverse kinematics, basic framework comprises motion capture module 1, action recognition module 2, dio Output Modules and GPRS alarm module 4.By Kinect camera captured at jobsite human action, use inverse kinematics software to carry out action recognition, characteristic parameter is acted on virtual human body, find that the behavior of stealing of doubtful car is reported to the police at once, alarm signal is outputted to loudspeaker 3 by dio Output Modules.
Described motion capture module mainly applies Kinect somatosensory device, draws the 3 d pose of human body by analyzing skeleton structure position model.The color image data of pedestrian's posture in garage, skeleton data and depth image data are obtained by colour imagery shot, infrared camera and image acquisition device in Kinect somatosensory device; Be 20 D coordinates value of joint in system spatial coordinates system in skeleton by the color image data of acquisition, skeleton data and depth image data transformations; Human body skeleton pattern can be expressed as level (tree-like) structure about node, represents a joint (or the directly actuated local coordinate system in joint), as shown in Figure 4 with each node.Introduce set membership between node, all nodes have common ancestors to be called root node, and with buttocks central authorities for root node, each node except root node has father node.Each node is placed relative to father node, represents, utilize tree structure with the conversion of a local, and a positive movement can be reduced to and be realized by recursive traversal tree construction.The degree of freedom of human skeleton model, for determining the independent variable number required for this configuration state completely, under the mutual standalone case in each joint, can be expressed as articulate degree of freedom summation.
Described action recognition module as shown in Figure 3, mainly utilizes turning axle representation rotation to be expressed as the vectorial N of 1 × 3.Rotating vector N represent around turning axle direction a anglec of rotation θ, wherein θ=|| a|| is the mould of N, a=[a
x, a
y, a
z]
tfor normalization sense of rotation, meet N=θ a.Corresponding rotation matrix R
n(3 × 3) can draw with following Rodrigues formulae discovery:
To given two vectorial a and b, definition directly rotates to be a rotation around axle c=a × b, and turning axle c is perpendicular to a and b, and a is rotated to the direction of vectorial b by rotation results.Its rotation matrix can calculate R (a, b)=R by turning axle representation
a × b(acos (a
tb)).
In order to recover the 3 d pose of human body, be that model is analyzed with skeleton structure.Matrix parameter determines the attitude of human body, adopts tree structure to represent human skeleton, and each joint of human body is a node in tree.A local coordinate system is adhered in each joint, and this coordinate system can rotate around this articulation point according to the rotation parameter in joint.Joint parameter comprises a translation vector T (1 × 3) and a rotating vector N (1 × 3).Utilize N and T can calculate the coordinate system of this articulation point and the relative rigid body translation of between father node coordinate system
Translation vector T illustrates the translation of this joint relative to father's articulation point, and the bone length size measured by color image data, skeleton data and depth image data etc. is determined, in tracing process, bone length is constant, and translation T is fixed value.Wherein bone length value mainly comprises 8: the bones of the body-neck, neck-shoulder, shoulder-elbow, the bones of the body-hip, hip-knee, neck-crown, elbow-hand, knee-foot, their size determines T value, and Li Heng (Xian Electronics Science and Technology University) studies in a literary composition in the bone recognition system based on Kinect bone following function and uses the euclidean distance classifier improved to carry out detailed elaboration to the identification of bone length by depth image and Kinect SDK.
Human body degree of freedom is the parameter that can independently change in skeleton.In the mutual independently situation in each joint of hypothesis, equal articulate degree of freedom summation.For picture elbow, this revolute joint of knee, we distribute 1 degree of freedom, and turning axle is fixed as x-axis.For ball groove joint models such as picture shoulder, hips, turning axle can be positioned at any direction of sphere, adds 1 degree of freedom pivoted, and is divided into and joins 3 degree of freedom.For neck joint, then distribute 2 degree of freedom, restriction turning axle is positioned at z=0 plane.Root node is positioned at the hip of health, its state modulator human skeleton position in world coordinate system, has 3 and rotates 3 translations, totally 6 degree of freedom.
Above translation rotation parameter is applied in the manikin set up in advance, draw the motion characteristic of simulation human body, with constantly perfect theft data characteristics storehouse carry out right, if similarity reaches h (danger coefficient, can change according to occasion and strick precaution degree), first cause the warning of control center's certain hour, then according to the accuracy of identification data, operating personnel can be had to cut off and report to the police or connect on-the-spot speaker unit.Along with the degree of perfection of database, when identification degree reaches
time more than (can according to strick precaution degree artificial regulable control), personnel control mode can be omitted, directly turn to on-site speaker device.
Described theft data characteristics storehouse is by artificially simulation or existing robber garage are that scene is carried out constantly perfect, the positive sample of not open close mistake and the continuous Real time identification of negative sample are optimized with identifying, for discharge robber unblank posture with adopt key unlocking posture to obscure may, adopt the identification optimization method based on hidden Markov model, Division identification between the action of solution mutual part overlap or similarity, this algorithm knows on method for distinguishing in independence before, candidate actions (is such as used and cuts short lock instrument, trial is repeatedly unblanked key, uncaging time is long) sequence adds restriction in sequential, specific algorithm application refers in research one literary composition of Hong Chen (Shanghai Communications University) 3 D human body action recognition in interactive system about action identification method, and tax weights are carried out to the important difference parameter of inverse kinematics feedback, and carry out positive and negative sample simulation to guarantee its accuracy identified.
The vehicle monitoring burglary-resisting system based on inverse kinematics of the present embodiment, comprise motion capture module 1, action recognition module 2, dio Output Modules 3, GPRS alarm module 4, motion capture module 1 mainly contains the PC that Kinect video camera carrys out capture site, action recognition module 2 has an automatic control, and system schematic is shown in accompanying drawing 1.Wherein the corner of Kinect video camera can be set by the watchdog routine on PC and control, and changes certain angle at set intervals to obtain wider monitoring visual field.The interface that the SDK (Software Development Kit) Kinect SDK that watchdog routine on PC contains Kinect provides, can obtain Skeleton Frame data stream that Kinect hardware transport comes and extract the x, y, z coordinate of each articulation point in human skeleton model.Dio Output Modules 3 is mainly made up of loudspeaker, is connected with PC, is controlled by the watchdog routine on PC, and give the alarm when " stealing the doubtful action of car " recognizing the sound alerting signal is sent to Security Personnel by GPRS warning device 4.
Claims (8)
1. based on a vehicle monitoring theft preventing method for inverse kinematics, it is characterized in that: described method comprises the steps:
Step 1, the Kinect video camera that bicycle shed is installed catches the action of human body in the visual field in real time, and the data of acquisition are transferred to coupled PC;
Step 2, PC obtains Skeleton Frame data stream by the interface that inverse kinematics software Kinect SDK provides and extracts the x, y, z coordinate data of each articulation point of human skeleton model;
Step 3, calculates the characteristic parameter between the direction vector of human body and position, by the virtual human model of restriction on the parameters in control inerface by the x, y, z coordinate of each articulation point of manikin;
Step 4, virtual human body motion characteristic and default " doubtful motion characteristic storehouse stolen by car " are carried out matching ratio comparatively, if when the difference parameter of virtual human body motion characteristic weighting reaches threshold value h with the similarity presetting the car stealing data characteristics storehouse and steal doubtful motion characteristic, be then that doubtful action stolen by car by preliminary judgement;
Step 5, monitors warning program and starts, carry out Real-time Alarm when human action is identified as " doubtful action stolen by car ".
2. as claimed in claim 1 based on the vehicle monitoring theft preventing method of inverse kinematics, it is characterized in that: in described step 5, first the dio Output Modules causing control center to be connected with PC gives the alarm the sound, then according to the accuracy of identification data, cut off by operating personnel and report to the police or connect on-the-spot speaker unit and report to the police.
3. realize, as claimed in claim 1 based on a burglary-resisting system for the vehicle monitoring theft preventing method of inverse kinematics, it is characterized in that: described burglary-resisting system comprises Kinect camera, motion capture module, action recognition module and alarm module,
Described motion capture module, receives the data of Kinect camera, and application Kinect somatosensory device, draws the 3 d pose of human body by analyzing skeleton structure position model;
Described action recognition module, utilizes turning axle representation rotation to be expressed as the vectorial N of 1 × 3, rotating vector N represent around turning axle direction a anglec of rotation θ, wherein θ=|| a|| is the mould of N, a=[a
x, a
y, a
z]
tfor normalization sense of rotation, meet N=θ a, corresponding rotation matrix R
n(3 × 3) draw with Rodrigues formulae discovery;
The root node of described human body degree of freedom is positioned at the hip of health, its state modulator human skeleton position in world coordinate system, has 3 and rotates 3 translations, totally 6 degree of freedom; Translation rotation parameter is applied in the manikin set up in advance, draws the motion characteristic of simulation human body, carries out right, if when similarity reaches threshold value h, be tentatively decided to be car and steal doubtful action with the theft data characteristics storehouse of presetting;
Described alarm module, monitors warning program and starts, carry out Real-time Alarm when human action is identified as " doubtful action stolen by car ".
4. burglary-resisting system as claimed in claim 3, is characterized in that: described alarm module comprises dio Output Modules and GPRS alarm module.
5. the burglary-resisting system as described in claim 3 or 4, it is characterized in that: in described motion capture module, obtain the color image data of pedestrian's posture in garage, skeleton data and depth image data by colour imagery shot, infrared camera and image acquisition device in Kinect somatosensory device; Be 20 D coordinates value of joint in system spatial coordinates system in skeleton by the color image data of acquisition, skeleton data and depth image data transformations; Human body skeleton pattern can be expressed as a tree structure about node, represents a joint with each node.
6. the burglary-resisting system as described in claim 3 or 4, is characterized in that: in described action recognition module, and human body degree of freedom is the parameter that can independently change in skeleton, in the mutual independently situation in each joint of hypothesis, equals articulate degree of freedom summation; For picture elbow, knee revolute joint, distribute 1 degree of freedom, turning axle is fixed as x-axis; For picture shoulder, hip ball groove joint model, turning axle is positioned at any direction of sphere, adds 1 degree of freedom pivoted, is divided into and joins 3 degree of freedom, for neck joint, then distributes 2 degree of freedom, and restriction turning axle is positioned at z=0 plane.
7. the burglary-resisting system as described in claim 3 or 4, it is characterized in that: in described action recognition module, described theft data characteristics storehouse is that scene is carried out constantly perfect by artificial simulation or existing robber garage, and the positive sample of not open close mistake and the continuous Real time identification of negative sample are optimized with identifying.
8. burglary-resisting system as claimed in claim 7, is characterized in that: in described theft data characteristics storehouse, adopt the identification optimization method based on hidden Markov model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410525013.0A CN104298353A (en) | 2014-10-08 | 2014-10-08 | Inverse kinematics based vehicle monitoring and burglary preventing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410525013.0A CN104298353A (en) | 2014-10-08 | 2014-10-08 | Inverse kinematics based vehicle monitoring and burglary preventing method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104298353A true CN104298353A (en) | 2015-01-21 |
Family
ID=52318115
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410525013.0A Pending CN104298353A (en) | 2014-10-08 | 2014-10-08 | Inverse kinematics based vehicle monitoring and burglary preventing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104298353A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105096696A (en) * | 2015-07-31 | 2015-11-25 | 努比亚技术有限公司 | Sign language translation apparatus and method based on intelligent bracelet |
| CN106485748A (en) * | 2015-08-25 | 2017-03-08 | 达索系统美国公司 | Method and system for the vision measurement of Digital human model |
| CN106564466A (en) * | 2015-10-12 | 2017-04-19 | 安徽大学 | Automobile antitheft alarm system based on behavior detection |
| CN106652291A (en) * | 2016-12-09 | 2017-05-10 | 华南理工大学 | Indoor simple monitoring and alarming system and method based on Kinect |
| CN107392957A (en) * | 2017-07-19 | 2017-11-24 | 杭州中赛实业有限公司 | A kind of children's garment fitting method and its children's garment based on body-sensing technology |
| CN108295469A (en) * | 2017-12-04 | 2018-07-20 | 成都思悟革科技有限公司 | Game visual angle effect method based on motion capture technology |
| CN109544606A (en) * | 2018-11-02 | 2019-03-29 | 山东大学 | Fast automatic method for registering and system based on multiple Kinect |
| CN109961473A (en) * | 2017-12-25 | 2019-07-02 | 深圳超多维科技有限公司 | Eyes localization method and device, electronic equipment and computer readable storage medium |
| WO2020000396A1 (en) * | 2018-06-29 | 2020-01-02 | Baidu.Com Times Technology (Beijing) Co., Ltd. | Theft proof techniques for autonomous driving vehicles used for transporting goods |
| CN111008601A (en) * | 2019-12-06 | 2020-04-14 | 江西洪都航空工业集团有限责任公司 | Fighting detection method based on video |
| CN111669652A (en) * | 2020-06-23 | 2020-09-15 | 深圳创维－Rgb电子有限公司 | Video monitoring and warning method and television |
| CN113807212A (en) * | 2021-08-31 | 2021-12-17 | 无锡瑞鼎电力科技有限公司 | Well lid anti-theft method and device based on image recognition |
| US11615648B2 (en) | 2021-05-28 | 2023-03-28 | Sportsbox.ai Inc. | Practice drill-related features using quantitative, biomechanical-based analysis |
| US11620783B2 (en) | 2021-05-27 | 2023-04-04 | Ai Thinktank Llc | 3D avatar generation and robotic limbs using biomechanical analysis |
| CN116328276A (en) * | 2021-12-22 | 2023-06-27 | 成都拟合未来科技有限公司 | Gesture interaction method, system, device and medium based on body building device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013092876A (en) * | 2011-10-25 | 2013-05-16 | Nippon Telegr & Teleph Corp <Ntt> | Attitude estimation device, attitude estimation method, and attitude estimation program |
| CN203325215U (en) * | 2013-06-21 | 2013-12-04 | 浙江理工大学 | Pedestrian red light running detection system |
| CN104038738A (en) * | 2014-06-04 | 2014-09-10 | 东北大学 | Intelligent monitoring system and intelligent monitoring method for extracting coordinates of human body joint |
-
2014
- 2014-10-08 CN CN201410525013.0A patent/CN104298353A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013092876A (en) * | 2011-10-25 | 2013-05-16 | Nippon Telegr & Teleph Corp <Ntt> | Attitude estimation device, attitude estimation method, and attitude estimation program |
| CN203325215U (en) * | 2013-06-21 | 2013-12-04 | 浙江理工大学 | Pedestrian red light running detection system |
| CN104038738A (en) * | 2014-06-04 | 2014-09-10 | 东北大学 | Intelligent monitoring system and intelligent monitoring method for extracting coordinates of human body joint |
Non-Patent Citations (1)
| Title |
|---|
| 洪琛: "交互式系统中三维人体动作识别的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105096696A (en) * | 2015-07-31 | 2015-11-25 | 努比亚技术有限公司 | Sign language translation apparatus and method based on intelligent bracelet |
| CN106485748A (en) * | 2015-08-25 | 2017-03-08 | 达索系统美国公司 | Method and system for the vision measurement of Digital human model |
| CN106485748B (en) * | 2015-08-25 | 2022-06-10 | 达索系统美国公司 | Method and system for visual measurement of digital mannequins |
| CN106564466A (en) * | 2015-10-12 | 2017-04-19 | 安徽大学 | Automobile antitheft alarm system based on behavior detection |
| CN106652291A (en) * | 2016-12-09 | 2017-05-10 | 华南理工大学 | Indoor simple monitoring and alarming system and method based on Kinect |
| CN107392957A (en) * | 2017-07-19 | 2017-11-24 | 杭州中赛实业有限公司 | A kind of children's garment fitting method and its children's garment based on body-sensing technology |
| CN107392957B (en) * | 2017-07-19 | 2021-06-08 | 杭州中赛实业有限公司 | Children dress fitting method based on somatosensory technology and children dress thereof |
| CN108295469B (en) * | 2017-12-04 | 2021-03-26 | 成都思悟革科技有限公司 | Game visual angle conversion method based on motion capture technology |
| CN108295469A (en) * | 2017-12-04 | 2018-07-20 | 成都思悟革科技有限公司 | Game visual angle effect method based on motion capture technology |
| CN109961473A (en) * | 2017-12-25 | 2019-07-02 | 深圳超多维科技有限公司 | Eyes localization method and device, electronic equipment and computer readable storage medium |
| WO2020000396A1 (en) * | 2018-06-29 | 2020-01-02 | Baidu.Com Times Technology (Beijing) Co., Ltd. | Theft proof techniques for autonomous driving vehicles used for transporting goods |
| CN111052193A (en) * | 2018-06-29 | 2020-04-21 | 百度时代网络技术（北京）有限公司 | Anti-theft technique for autonomous vehicles for transporting goods |
| CN109544606A (en) * | 2018-11-02 | 2019-03-29 | 山东大学 | Fast automatic method for registering and system based on multiple Kinect |
| CN109544606B (en) * | 2018-11-02 | 2022-02-15 | 山东大学 | Rapid automatic registration method and system based on multiple Kinects |
| CN111008601A (en) * | 2019-12-06 | 2020-04-14 | 江西洪都航空工业集团有限责任公司 | Fighting detection method based on video |
| CN111669652A (en) * | 2020-06-23 | 2020-09-15 | 深圳创维－Rgb电子有限公司 | Video monitoring and warning method and television |
| US11620783B2 (en) | 2021-05-27 | 2023-04-04 | Ai Thinktank Llc | 3D avatar generation and robotic limbs using biomechanical analysis |
| US11615648B2 (en) | 2021-05-28 | 2023-03-28 | Sportsbox.ai Inc. | Practice drill-related features using quantitative, biomechanical-based analysis |
| US11620858B2 (en) | 2021-05-28 | 2023-04-04 | Sportsbox.ai Inc. | Object fitting using quantitative biomechanical-based analysis |
| US11640725B2 (en) | 2021-05-28 | 2023-05-02 | Sportsbox.ai Inc. | Quantitative, biomechanical-based analysis with outcomes and context |
| US11935330B2 (en) | 2021-05-28 | 2024-03-19 | Sportsbox.ai Inc. | Object fitting using quantitative biomechanical-based analysis |
| US11941916B2 (en) | 2021-05-28 | 2024-03-26 | Sportsbox.ai Inc. | Practice drill-related features using quantitative, biomechanical-based analysis |
| CN113807212A (en) * | 2021-08-31 | 2021-12-17 | 无锡瑞鼎电力科技有限公司 | Well lid anti-theft method and device based on image recognition |
| CN116328276A (en) * | 2021-12-22 | 2023-06-27 | 成都拟合未来科技有限公司 | Gesture interaction method, system, device and medium based on body building device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104298353A (en) | Inverse kinematics based vehicle monitoring and burglary preventing method and system | |
| CN109819208B (en) | Intensive population security monitoring management method based on artificial intelligence dynamic monitoring | |
| CN105856230B (en) | A kind of ORB key frames closed loop detection SLAM methods for improving robot pose uniformity | |
| CN110135249B (en) | Human behavior identification method based on time attention mechanism and LSTM (least Square TM) | |
| Hu et al. | A survey on visual surveillance of object motion and behaviors | |
| CN112200106A (en) | Cross-camera pedestrian re-identification and tracking method | |
| US20220180534A1 (en) | Pedestrian tracking method, computing device, pedestrian tracking system and storage medium | |
| CN110008913A (en) | The pedestrian's recognition methods again merged based on Attitude estimation with viewpoint mechanism | |
| CN107397658B (en) | Multi-scale full-convolution network and visual blind guiding method and device | |
| CN110334701A (en) | Collecting method based on deep learning and multi-vision visual under the twin environment of number | |
| CN105059190A (en) | Vision-based automobile door-opening bump early-warning device and method | |
| CN112733707A (en) | Pedestrian re-identification method based on deep learning | |
| CN108346265B (en) | A kind of risk object identification alarm method | |
| Wang et al. | Deep learning based target detection algorithm for motion capture applications | |
| Wu et al. | A stepwise minimum spanning tree matching method for registering vehicle-borne and backpack LiDAR point clouds | |
| Chang et al. | Vision-aware target recognition toward autonomous robot by Kinect sensors | |
| Saif et al. | Crowd density estimation from autonomous drones using deep learning: challenges and applications | |
| Tarko et al. | Tscan: Stationary lidar for traffic and safety studies—object detection and tracking | |
| Datondji et al. | Rotation and translation estimation for a wide baseline fisheye-stereo at crossroads based on traffic flow analysis | |
| CN115979250B (en) | Positioning method based on UWB module, semantic map and visual information | |
| Özbek et al. | Subjective analysis of social distance monitoring using YOLO v3 architecture andcrowd tracking system | |
| Bai et al. | Cyber mobility mirror for enabling cooperative driving automation: A co-simulation platform | |
| CN110276379A (en) | A kind of the condition of a disaster information rapid extracting method based on video image analysis | |
| CN115761618A (en) | Key site security monitoring image identification method | |
| Li et al. | PointLAE: A Point Cloud Semantic Segmentation Neural Network via Multifeature Aggregation for Large-Scale Application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150121 |