CN108764203A - A kind of pedestrian's quantitative analysis and display systems towards urban planning - Google Patents
A kind of pedestrian's quantitative analysis and display systems towards urban planning Download PDFInfo
- Publication number
- CN108764203A CN108764203A CN201810574417.7A CN201810574417A CN108764203A CN 108764203 A CN108764203 A CN 108764203A CN 201810574417 A CN201810574417 A CN 201810574417A CN 108764203 A CN108764203 A CN 108764203A
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- data
- corresponding time
- analysis
- flow
- 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
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The pedestrian's quantitative analysis and display systems that the present invention provides a kind of towards urban planning, the headend equipment for being mainly combined video acquisition module and embedding assembly platform is as pedestrian's analysis system hardware platform designed by this paper.This method includes:Combining target detects and pedestrian's flow analysis of tracking is trained using MobileNet-SSD networks, obtains pedestrian target detection model.The data set that the convolutional neural networks training test of bridging is established herein is improved, face gender and Analysis of age is obtained, is finally shown at the ends Web.The present invention has given full play to the superiority of embedded platform, avoids the limitation of traditional artificial extraction feature, is designed in the system of embedded platform using the realization of the method for the present invention.It is freed in the huge workload of video big data analysis to Urban Planner, and more data supportings is provided for Urban Planner.
Description
Technical field
The present invention relates to pedestrian's quantitative analysis problems in urban planning, especially obtain pedestrian's flow in video, are detained
Pedestrians' dependent quantization data such as time, motion track, gender and age realize that pedestrian's quantitative analysis and display systems are advised in city
Draw the research work in project.
Background technology
As Re-search on Urbanization persistently promotes, novel urbanization is the urbanization taking human as core.It is people-oriented building
Urbanization process in, analysis mankind's activity data be to realize the basic skills of the urban planning taking human as core.And the mankind live
Dynamic data are largely divided into three kinds of classifications:Group's quantitative information (pedestrian's flow, residence time, motion track, gender, age etc.);
Crowd behaviour (individual and group behavior) information;Crowd's mood (individual and group's mood) information.
Group's quantitative analysis is mainly used at present, rough group's Population size estimation is realized based on Density Detection algorithm,
But the analysis pedestrian's individual information that can not be refined.Pedestrian's flow can be used for the residence time commenting with pedestrian in group's quantitative analysis
Estimate Street vitality, and Street vitality is mainly for assessment of the quality (street design reasonability) in street.Group's quantitative analysis is neutral
It is not mainly used for the friendly for analyzing street for different sexes or all ages and classes stratum pedestrian with the age.Wherein pedestrian stream
Amount, residence time, motion track direction, gender and age are URBAN PLANNING STUDY " avenue vitality assessment ", " space
Maximally related five kinds of data in the hot fields such as quality evaluation ".
Video data can be used for obtaining the data such as flow of the people, gender and age.It is compared in existing research and is adopted simultaneously
Big data is enriched with information completely, dimension and the video big data of the adjustable three advantages of granularity can be public to city
The mankind's activity scene of highly dynamicization carries out fining description in space, has huge Research Prospects.With embedded flat
Platform performance gradually increases, and embedded platform has the features such as portability, low-power consumption, and more and more persons select in insertion
Video intelligent analysis is carried out on formula platform.Therefore it is urban planning field profit to carry out video pedestrian analysis based on embedded system
The inevitable development trend of data acquisition and analysis is carried out with video.
Invention content
It is realized the object of the present invention is to provide a kind of method and realizes pedestrian's quantitative analysis and displaying system on embedded platform
System to obtain the useful information in video big data, and does the assessment of urban vitality convenient for Urban Planner
Go out more correct decision.
For convenience of explanation, it is firstly introduced into following concept:
Convolutional neural networks (CNN):It is designed by the inspiration of optic nerve mechanism, is designed for identification two-dimensional shapes
A kind of multilayer perceptron, this network structure has height to the deformations of translation, proportional zoom, inclination or other forms
Invariance.
Caffe frames:It is a deep learning frame of increasing income being widely used, the network in Caffe is all oriented nothing
The set of ring figure, data and its derivative are in the form of blobs in interflow.
AngularJS:A outstanding front end JS frames, core is the most:MVW(Model-View-Whatever),
Modularization, the binding of automation bi-directional data, semantization label, dependence injection etc..
Django:The Web application frameworks of one open source code, are write as by Python.The framework mode of MTV is used,
That is model M, template T and view V.
MTCNN algorithms:It is a kind of fast convolution neural network algorithm of three ranks cascade mode.There are three the algorithm is total
Stage:First stage quickly generates a large amount of candidate frame by shallow-layer CNN (Convolutional Neural Networks);
Second stage refines candidate frame using more complicated CNN, abandons the largely candidate frame without face;Final stage is using more
Powerful CNN realizes the candidate frame for selecting final face, and exports five people's face portions key point position.
TensorRT:It is that Nvidia disposes deep learning application program in release in 2016 is dedicated for production environment
A kind of high-performance ANN Reasoning engine.
Pedestrian's data set:These data sets are finally divided into training set and test set, and wherein training set is used for the training stage
Training data and test data, test set be used for finally training complete MobileNet-SSD networks and parameter model into
Row test.
The technical proposal of the invention is realized in this way:
In a first aspect, the pedestrian's quantitative analysis that an embodiment of the present invention provides a kind of towards urban planning and display systems
Equipment, the equipment include:Camera, embedded processing platform, data storage server and the ends Web;
The camera is connected with embedded processing platform;The camera, the video for acquiring street pedestrian;Institute
Embedded processing platform is stated, the video of the pedestrian for being acquired by the camera carries out phase under Ubuntu16.04 systems
It closes and calculates;The data storage server, the data for being calculated by the embedded processing platform are stored in server;
The ends Web, the data for being stored by data storage server eventually pass through processing and are shown at the ends Web.
In conjunction with system hardware Platform Designing, an embodiment of the present invention provides a kind of possible embodiments, by video acquisition
The headend equipment that module and embedding assembly platform are combined is as pedestrian's analysis system hardware platform designed by this paper;
Ubuntu systems are run using Jetson TX embedded platforms, are based partially on the program of the ends PC exploitation without carrying out
Transplanting can directly be run in Jetson TX2 platforms;
It is tied up and is installed suitable for NVIDIA Jetson embedded platforms using the JetPack integrated software package bundles provided
All exploitation software tools, TX2 platforms run Caffe frames.
Second aspect, is incorporated in embedding assembly platform object detection task, and the present invention provides a kind of possible embodiment party
Formula, wherein pedestrian's flow analysis of the target detection and tracking is used for:
Pedestrian detection is carried out to video first, then judges whether the pedestrian target of detection and current tracking target are homogeneous
Together, it will be added in pedestrian tracking sequence from the different detection pedestrian target of tracking target if there is differing target then, and
Update pedestrian's flow variable;Update pedestrian flow is jumped directly to if pedestrian target all same in next step;Next judge to regard
Whether frequency stream terminates, the pedestrian detection before being continued to execute if not and operation later;If video flowing terminates, note
Picture recording closes quantized data.
The third aspect, is incorporated in pedestrian's flow analysis of target detection and tracking, and the present invention provides a kind of possible implementation
Mode, wherein pedestrian's attributive analysis of the face is used for:
Realize MTCNN in TX2 first with pedestrian's face that MTCNN algorithm detecting and trackings arrive, while using TensorRT
Optimization transplanting development on platform.Then gender analysis is carried out to the face detected, since face quantity can excessively influence to locate
Efficiency is managed, therefore improves the convolutional neural networks based on bridging, while ensureing speed, it is promoted and differentiates effect;Exist simultaneously
It is trained and verifies on the gender data collection established herein;Then age bracket differentiation processing is carried out to the face detected, together
When the age segment data set established herein on trained and verified, finally carrying out systematic entirety to TX2 platforms can survey
Examination.
Fourth aspect, server-side carry out visualization data by Web page and show that the present invention provides a kind of possible implementation
Mode, wherein Web end systems use B/S frameworks, the client that browser is used as user mainly graphically to open up
Show video line personal data treated quantitative information;
This paper systems using AngularJS as front end frame, using Django MTV as rearward end frame, Django editions
This is 1.8, using MySQL database for storing related data.
To keep the above objects, features and advantages of the present invention more obvious and easy to understand, hereafter coordinate appended attached drawing, makees detailed
It is described as follows.
Description of the drawings
Fig. 1 is the system the general frame of the present invention;
Fig. 2 is the system entirety software architecture of the present invention;
A kind of pedestrian detection and track overall plan flow chart that Fig. 3 is carried by the embodiment of the present invention;
Fig. 4 is detected and pedestrian tracking flow chart by a kind of pedestrian target that the embodiment of the present invention carries;
Fig. 5 is the database design drawing of the present invention;
A kind of pedestrian information chart that Fig. 6 is carried by the embodiment of the present invention;
A kind of face character chart that Fig. 7 is carried by the embodiment of the present invention;
Specific implementation mode
Below by embodiment, the present invention is described in further detail, it is necessary to, it is noted that embodiment below
It is served only for that the present invention is described further, should not be understood as limiting the scope of the invention, fields technology is ripe
Personnel are known according to foregoing invention content, some nonessential modifications and adaptations are made to the present invention and are embodied, should still be belonged to
In protection scope of the present invention.
The pedestrian's quantitative analysis and display systems that an embodiment of the present invention provides a kind of towards urban planning, below by reality
Example is applied to be described.
In system hardware platform, the headend equipment that whole system camera and embedded processing platform are combined is as this
Pedestrian's analysis system hardware platform designed by text, using NVIDIA Jetson TX2 embedded processing platforms, Jetson
TX2 platforms run Ubuntu systems, by the way that TX2 platforms to be tightly fixed to together with video camera, while using network deconcentrator
TX2 platforms are connected with camera, only need a cable that can make TX2 platforms while camera data can be obtained in this way
And access server end.
Pedestrian detection is carried out using MobileNet-SSD algorithm of target detection to video first, then judges the row of detection
People's target and current tracking target whether all same, if there is differing target then by the detection row different from tracking target
People's target is added in pedestrian tracking sequence, and updates pedestrian's flow variable;It is jumped directly to more if pedestrian target all same
Newline flow of the people is in next step;Next judge whether video flowing terminates, pedestrian detection before being continued to execute if not and
Operation later;If video flowing terminates, dependent quantization data are recorded.
Pedestrian's attributive analysis that next step face is carried out after the completion of pedestrian detection and pedestrian tracking, first with MTCNN
Pedestrian's face that algorithm detecting and tracking arrives, while realizing optimization transplanting developments of the MTCNN on TX2 platforms using TensorRT.
Then gender analysis is carried out to the face detected since face quantity can excessively influence treatment effeciency to change
Into the convolutional neural networks based on bridging, while ensureing speed, promotes it and differentiate effect;Urbanists are main
And the non-interesting specific age, and focus more on its age bracket classification, as infant, teenager, the middle age and it is old these four
Classification.
The development deployment platform of Web end systems is 64 bit manipulation systems of Ubuntu16.04, the Integrated Development ring of use
Border (Integrated Development Environment, IDE) is Pycharm and Sublime Text3, Development Framework
For MTV software architectures.
Web end systems are mainly used for pedestrian's flow, residence time, motion track, gender and the age letter of displaying processing
Cease the quantized data of gained;Tables of data is used to record the information of all concern street cameras, including No. id of camera,
Manage position longitude and latitude, residing street name and the time for adding the concern street point;Pedestrian information table include processing time,
All pedestrian's flows in the corresponding time, track moves to left that pedestrian's flow, track moves to right pedestrian stream in the corresponding time in the corresponding time
Amount, in the corresponding time quickly through pedestrian's flow, middling speed is led at a slow speed by pedestrian's flow and in the corresponding time in the corresponding time
Space flow of the people;Face character table includes male's quantity, women quantity, correspondence in the corresponding time in processing time, corresponding time
Infant's quantity, teenager's quantity, middle aged quantity, old quantity in the corresponding time in the corresponding time in the corresponding time in time.
All camera position information in the ends Web and pedestrian's dependent quantization information are all made of analogue data and carry out effect displaying,
Web end systems use charts of the visualization library ECharts for pedestrian information data of increasing income developed based on JavaScript to paint
System realizes the function of pedestrian information data displaying after point processing in street of interest.
Claims (7)
1. a kind of pedestrian's quantitative analysis and display systems towards urban planning, which is characterized in that including:It is camera, embedded
Processing platform, data storage server and the ends Web;The camera is connected with embedded processing platform;
The camera, the video for acquiring street pedestrian;
The embedded processing platform, the video of the pedestrian for being acquired by the camera is under Ubuntu16.04 systems
Carry out correlation computations;
The data storage server, the data for being calculated by the embedded processing platform are stored in server;
The ends Web, the data for being stored by data storage server eventually pass through processing and are shown at the ends Web.
2. equipment according to claim 1, which is characterized in that the embedded processing platform includes:NVIDIA Jetson
TX2;The embedded processing platform is specifically used for:
Target detection based on deep learning and target following are completed to collected pedestrian's video, while judging the movement of pedestrian
Course bearing and calculating residence time, and relevant information is preserved into database;
The analysis design of individual attribute based on face is completed to pedestrian detection, the data of pedestrian tracking and is realized.
3. equipment according to claim 1, which is characterized in that the data storage server is specifically used for:
In the server to the data storage after quantifying described in right 2, including following information, storage data information table, Hang Renliang
Change table, face character table.
4. according to data information table described in right 3, it is characterised in that:
Tables of data be used for record it is all concern street cameras information, including No. id of camera, geographical location longitude and latitude, institute
The street name at place and the time for adding the concern street point.
5. quantifying table according to pedestrian described in right 3, it is characterised in that:
For recording the relevant information after current street camera quantification treatment in database, in tables of data comprising processing time,
All pedestrian's flows in the corresponding time, in the corresponding time track move to left pedestrian's flow, in the corresponding time track move to right pedestrian's flow,
Quickly through pedestrian's flow, middling speed passes through at a slow speed row by pedestrian's flow and in the corresponding time in the corresponding time in the corresponding time
Flow of the people.
6. according to face character table described in right 3, it is characterised in that:
Include male's quantity, women quantity, baby children in the corresponding time in the corresponding time in processing time, corresponding time in tables of data
Youngster's quantity, teenager's quantity, middle aged quantity, old quantity in the corresponding time in the corresponding time in the corresponding time.
7. equipment according to claim 1, which is characterized in that the ends Web are specifically used for:
Show the quantized data of the pedestrian's flow handled, residence time, motion track, gender and age information.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810574417.7A CN108764203A (en) | 2018-06-06 | 2018-06-06 | A kind of pedestrian's quantitative analysis and display systems towards urban planning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810574417.7A CN108764203A (en) | 2018-06-06 | 2018-06-06 | A kind of pedestrian's quantitative analysis and display systems towards urban planning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108764203A true CN108764203A (en) | 2018-11-06 |
Family
ID=63999809
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810574417.7A Pending CN108764203A (en) | 2018-06-06 | 2018-06-06 | A kind of pedestrian's quantitative analysis and display systems towards urban planning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108764203A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109766848A (en) * | 2019-01-15 | 2019-05-17 | 四川大学 | A kind of pedestrian's eyesight status investigation and analysis method based on monitor video |
CN109948550A (en) * | 2019-03-20 | 2019-06-28 | 北京百分点信息科技有限公司 | A kind of wisdom railway station flow of the people monitoring system and method |
CN110175562A (en) * | 2019-05-24 | 2019-08-27 | 深圳职业技术学院 | A kind of gpu acceleration pedestrian detection algorithm based on parallel computation |
CN110351526A (en) * | 2019-07-23 | 2019-10-18 | 北京中安信合科技有限公司 | A kind of personnel's supervisory systems based on artificial intelligence technology |
CN112270381A (en) * | 2020-11-16 | 2021-01-26 | 电子科技大学 | People flow detection method based on deep learning |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN203301544U (en) * | 2013-06-05 | 2013-11-20 | 四川省那科通信有限责任公司 | Smart city monitoring system |
US20160171330A1 (en) * | 2014-12-15 | 2016-06-16 | Reflex Robotics, Inc. | Vision based real-time object tracking system for robotic gimbal control |
CN106529402A (en) * | 2016-09-27 | 2017-03-22 | 中国科学院自动化研究所 | Multi-task learning convolutional neural network-based face attribute analysis method |
CN107330372A (en) * | 2017-06-05 | 2017-11-07 | 四川大学 | A kind of crowd density based on video and the analysis method of unusual checking system |
CN107516136A (en) * | 2017-08-24 | 2017-12-26 | 北京鼎通远科技发展有限公司 | Power Line Inspection System |
CN107717989A (en) * | 2017-10-11 | 2018-02-23 | 深圳市普渡科技有限公司 | One kind uses intelligent distribution robot control system |
CN107818302A (en) * | 2017-10-20 | 2018-03-20 | 中国科学院光电技术研究所 | Non-rigid multiple dimensioned object detecting method based on convolutional neural networks |
-
2018
- 2018-06-06 CN CN201810574417.7A patent/CN108764203A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN203301544U (en) * | 2013-06-05 | 2013-11-20 | 四川省那科通信有限责任公司 | Smart city monitoring system |
US20160171330A1 (en) * | 2014-12-15 | 2016-06-16 | Reflex Robotics, Inc. | Vision based real-time object tracking system for robotic gimbal control |
CN106529402A (en) * | 2016-09-27 | 2017-03-22 | 中国科学院自动化研究所 | Multi-task learning convolutional neural network-based face attribute analysis method |
CN107330372A (en) * | 2017-06-05 | 2017-11-07 | 四川大学 | A kind of crowd density based on video and the analysis method of unusual checking system |
CN107516136A (en) * | 2017-08-24 | 2017-12-26 | 北京鼎通远科技发展有限公司 | Power Line Inspection System |
CN107717989A (en) * | 2017-10-11 | 2018-02-23 | 深圳市普渡科技有限公司 | One kind uses intelligent distribution robot control system |
CN107818302A (en) * | 2017-10-20 | 2018-03-20 | 中国科学院光电技术研究所 | Non-rigid multiple dimensioned object detecting method based on convolutional neural networks |
Non-Patent Citations (6)
Title |
---|
ROHIT PANDEY 等: "Real-time Egocentric Gesture Recognition on Mobile Head Mounted Display", 《HTTPS://ARXIV.ORG/ABS/1712.04961》 * |
冯帅 等: "基于Jetson TK1和深度卷积神经网络的行人检测", 《信息技术》 * |
张婷 等: "基于跨连卷积神经网络的性别分类模型", 《自动化学报》 * |
施一飞: "对使用TensorRT加速AI深度学习推断效率的探索", 《科技视界》 * |
曹诚 等: "城市量化研究中视频人流统计分析", 《计算机***应用》 * |
林智莘 等: "SAR舰船快速检测***设计", 《物联网技术》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109766848A (en) * | 2019-01-15 | 2019-05-17 | 四川大学 | A kind of pedestrian's eyesight status investigation and analysis method based on monitor video |
CN109766848B (en) * | 2019-01-15 | 2022-05-03 | 四川大学 | Pedestrian vision condition investigation and analysis method based on monitoring video |
CN109948550A (en) * | 2019-03-20 | 2019-06-28 | 北京百分点信息科技有限公司 | A kind of wisdom railway station flow of the people monitoring system and method |
CN110175562A (en) * | 2019-05-24 | 2019-08-27 | 深圳职业技术学院 | A kind of gpu acceleration pedestrian detection algorithm based on parallel computation |
CN110175562B (en) * | 2019-05-24 | 2024-01-16 | 深圳职业技术大学 | Parallel computing-based gpu acceleration pedestrian detection algorithm |
CN110351526A (en) * | 2019-07-23 | 2019-10-18 | 北京中安信合科技有限公司 | A kind of personnel's supervisory systems based on artificial intelligence technology |
CN112270381A (en) * | 2020-11-16 | 2021-01-26 | 电子科技大学 | People flow detection method based on deep learning |
CN112270381B (en) * | 2020-11-16 | 2022-06-03 | 电子科技大学 | People flow detection method based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108764203A (en) | A kind of pedestrian's quantitative analysis and display systems towards urban planning | |
Xiao et al. | A framework for quantitative analysis and differentiated marketing of tourism destination image based on visual content of photos | |
CN105354548B (en) | A kind of monitor video pedestrian recognition methods again based on ImageNet retrievals | |
CN110781765A (en) | Human body posture recognition method, device, equipment and storage medium | |
CN103988232B (en) | Motion manifold is used to improve images match | |
Bauder | Using GPS supported speed analysis to determine spatial visitor behaviour | |
CN107423398A (en) | Exchange method, device, storage medium and computer equipment | |
Lamas et al. | MonuMAI: Dataset, deep learning pipeline and citizen science based app for monumental heritage taxonomy and classification | |
Huai et al. | Using social media photos and computer vision to assess cultural ecosystem services and landscape features in urban parks | |
Heilbron et al. | Collecting and annotating human activities in web videos | |
CN114387499A (en) | Island coastal wetland waterfowl identification method, distribution query system and medium | |
Suzuki et al. | Enhancement of gross-motor action recognition for children by CNN with OpenPose | |
CN110378736B (en) | Method for evaluating experience satisfaction degree of tourists on natural resources through facial expression recognition | |
Padmaja et al. | A comparison on visual prediction models for MAMO (multi activity-multi object) recognition using deep learning | |
Sonnleitner et al. | Traffic measurement and congestion detection based on real-time highway video data | |
TW202240471A (en) | Methods, apparatuses, devices, and storage media for detecting target | |
Fei et al. | Flow-pose Net: An effective two-stream network for fall detection | |
Alharbey et al. | Human Faces Detection and Tracking for Crowd Management in Hajj and Umrah. | |
Mazzamuto et al. | Weakly supervised attended object detection using gaze data as annotations | |
CN117520662A (en) | Intelligent scenic spot guiding method and system based on positioning | |
Zhang et al. | Urban visual intelligence: Studying cities with AI and street-level imagery | |
Affes et al. | Real Time Detection and Tracking in Multi Speakers Video Conferencing | |
Sun et al. | Automatic building age prediction from street view images | |
Niu | A Lightweight Two-stream Fusion Deep Neural Network Based on ResNet Model for Sports Motion Image Recognition | |
Zhang et al. | Toward action recognition and assessment using SFAGCN and combinative regression model of spatiotemporal features |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181106 |
|
RJ01 | Rejection of invention patent application after publication |