CN110458025B - Target identification and positioning method based on binocular camera - Google Patents
Target identification and positioning method based on binocular camera Download PDFInfo
- Publication number
- CN110458025B CN110458025B CN201910625272.3A CN201910625272A CN110458025B CN 110458025 B CN110458025 B CN 110458025B CN 201910625272 A CN201910625272 A CN 201910625272A CN 110458025 B CN110458025 B CN 110458025B
- Authority
- CN
- China
- Prior art keywords
- target
- information
- picture
- personnel
- binocular camera
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06V40/172—Classification, e.g. identification
Abstract
The invention discloses a target identification and positioning method of a binocular camera. Carrying out classification learning based on a Convolutional Neural Network (CNN) by utilizing a face picture training database to obtain a classification model based on face recognition; and performing regression learning based on a Support Vector Machine (SVM) by using the depth information of the picture to obtain a distance regression function, and constructing the relation between the depth information of the picture and the distance between the depth information of the picture and the target distance camera. And shooting a target through a camera, and completing personnel target identification through a human face classification model by using the obtained human face picture. Meanwhile, the distance between the target and the camera is calculated by using the depth information of the shot picture, so that personnel positioning is realized. The method has the advantages of low recognition time overhead and high recognition precision.
Description
Technical Field
The invention relates to a target identification and positioning method of a binocular camera, and belongs to the technical field of positioning and navigation.
Background
In recent years, the demand for indoor location services has increased, and the development of indoor positioning technology has been promoted. Traditional satellite positioning system, like Global Positioning System (GPS), big dipper positioning system possess higher positioning accuracy in outdoor spacious environment, but satellite positioning signal receives to shelter from or disturb very easily, leads to satellite positioning system to fix a position inaccurate or even unable location in indoor environment. Therefore, in an indoor environment, the characteristics of no electromagnetic interference, environmental protection and the like are widely concerned in consideration of the advantages of image information.
The prior art includes: an LED visible light indoor positioning method based on image matching and a fingerprint database (patent number: CN 201610125773.1) uses a sift algorithm in an online stage, the online processing time is long, and the position estimation time overhead is large. The trained model is used in the on-line stage, so that the position estimation time cost is low, and the time required by positioning can be obviously prolonged.
Currently, a monocular camera or a binocular camera is used in the research of indoor positioning technology based on images at home and abroad. Among them, using binocular cameras has more advantages. Before stereoscopic vision, monocular vision has an absolute advantage in the field of identification and positioning tracking of people due to the advantages of high operation efficiency and small information amount. With the continuous development of stereoscopic vision, the defects that monocular vision cannot obtain image depth information and cannot accurately identify and position personnel targets are more obvious, binocular stereoscopic vision simulates human eyes through a binocular camera, parallax information of binocular images is extracted, and identification and positioning processing is further carried out according to the depth information and respective characteristics of objects in the images. And in the left and right binocular images acquired by the binocular camera, the horizontal distance between the central pixels of the two matching blocks is the parallax. The same parallax (i.e., the same color) represents that the object is at the same position from the camera. At present, with the development of hardware systems, particularly embedded systems, the advantages of binocular technology are more and more obvious in the field of image processing of moving objects. Therefore, the binocular vision technology is of great significance in application to intelligent traffic and video monitoring.
The information disclosed in this background section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art.
Disclosure of Invention
The invention aims to provide a binocular camera-based personnel identification and positioning method, so that the defects in the prior art are overcome.
In order to achieve the purpose, the invention provides a target identification and positioning method of a binocular camera, which comprises the following steps: step 1: carrying out classification learning based on a Convolutional Neural Network (CNN) by using a face picture training database to obtain a classification model based on face recognition;
step 2: carrying out regression learning based on a Support Vector Machine (SVM) by using the depth information of the picture to obtain a distance regression function, and constructing the relation between the depth information of the picture and the distance between the depth information of the picture and a target distance camera;
and step 3: shooting a target through a camera, and completing person target identification through a face classification model by using an obtained face picture;
and 4, step 4: meanwhile, the distance between the target and the camera is calculated by using the depth information of the shot picture, so that personnel positioning is realized.
The invention further defines the technical scheme as follows:
preferably, in the above technical solution, the target identification and positioning method of the binocular camera specifically includes:
step 1: shooting face image information of a person to be identified by using a binocular camera, and establishing a person target label and a face image database; performing offline classification learning on the personnel target labels and the face image database by using a convolutional neural network to obtain a personnel target identification classification model;
step 2: recording the position information of a shooting point, extracting the depth information of a shooting picture from a binocular camera shooting depth picture by utilizing an image processing technology in OpenCV, and establishing a position information and depth picture information database; training a position information and depth picture information database by using a support vector machine, and learning the corresponding relation between the position information and the depth picture information to obtain a regression model based on the position information;
and step 3: acquiring motion information of a target by using a binocular camera, and after the binocular camera shoots personnel pictures, inputting detected face images into the personnel target identification classification model in the step 1 through cutting processing by using a face detection algorithm to realize personnel identification;
and 4, step 4: and (3) when the system detects the target, automatically processing the acquired depth picture information of the target, and then inputting the depth picture information into the regression model based on the position information in the step (2) to obtain the specific position of the target from the camera, further obtain the position of the target and realize the positioning of personnel.
Preferably, step 1 and step 2 are off-line stages and step 3 and step 4 are on-line stages.
Preferably, the face detection algorithm is a Cascade Classifier algorithm.
The utility model provides a personnel discernment and positioning system based on binocular camera, includes personnel identification system and personnel positioning system, its characterized in that:
each of which in turn comprises two phases, an offline phase and an online phase.
(1) The identification process of the personnel comprises the following steps:
the function of the module is as follows: firstly, a face detection module is called, when a face is detected, a detected face image is cut and processed, then the cut face image is input into a trained person target recognition classification model, and the model can automatically recognize the detected person to obtain a recognition result. The convolutional neural network-based personnel target identification classification model is divided into two stages: an offline phase and an online phase.
An off-line stage: the binocular camera is used for shooting the face image information of the person to be recognized, and a (person target label, face picture) database is built. And (4) carrying out offline classification learning on the database (personnel target labels, human face pictures) by using a convolutional neural network to obtain a personnel target identification classification model.
An online stage: after the binocular camera shoots the picture of the person, a face detection algorithm (Cascade Classifier algorithm) is used for cutting the detected face image, and the person target recognition classification model is used for realizing the recognition of the person.
(2) The positioning process of the personnel comprises the following steps:
the personnel positioning regression model based on the support vector machine is divided into two stages: an offline phase and an online phase.
An off-line stage: and recording the position information of the shot point, extracting the depth information of the shot picture from the shot depth picture of the binocular camera, and establishing a (position information, depth picture information) database. And training the (position information and depth picture information) database by using a support vector machine, learning the corresponding relation between the position information and the depth picture information, and obtaining a regression model based on the position information.
An online stage: the method comprises the steps of collecting video information at an unknown position by using a binocular camera, processing a corresponding depth information picture after people are detected, carrying out normalization processing on pixel information of the depth information picture, and obtaining the specific position of a target from the camera by using a regression model of position information so as to obtain the position of the target.
Personnel discernment and positioning system based on binocular camera, including personnel identification system and personnel positioning system, its characterized in that: the hardware equipment of the invention mainly comprises: the device comprises an image acquisition device, an algorithm processing device and a display device.
An image acquisition device: the task uses a small foraging binocular camera to complete image acquisition, the model is S series (S1030-IR-120/MONO), and the camera parameters are as follows: replace Standard M12 lens, USB3.0 interface. The apparatus is shown in fig. 6:
FIG. 6: small-foraging binocular camera
An algorithm processing device: the collected image data needs further identification and positioning processing of personnel, and simulation experiments are all realized on a computer. The video processing apparatus is a computer, and its specific configuration: intel (R) Core (TM) i5-7300HQ CPU 2.50GHz with 7.89G memory.
A system display device: the result after the algorithm processing needs to be displayed and output, which is actually the display of the computer to accomplish this function.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention converts the problem of personnel target identification into the problem of classification based on the convolutional neural network. And in the online stage, the personnel target recognition classification model trained in the offline stage is utilized to realize the recognition of personnel. The method has the advantages of low recognition time overhead and high recognition precision.
2. The invention converts the personnel position estimation problem into a regression problem based on a support vector machine. And in the online stage, the target position estimation is realized by using the depth information of the image and a regression model of the position information. The method has the advantages of small position estimation time overhead and high positioning accuracy.
3. In the position estimation process of the personnel target, the invention realizes positioning by utilizing the depth information of the picture, thereby reducing the complexity of the image-based positioning method and simultaneously providing the positioning precision.
Drawings
FIG. 1 is an abstract of the specification of the present invention.
FIG. 2 is a flow chart of human target identification according to the present invention.
Fig. 3 is a flowchart of the convolutional neural network offline learning.
FIG. 4 is a flow chart of the present invention for estimating the position of a human target.
Fig. 5 is image depth information photographed by the binocular camera.
Fig. 6 shows the result of estimating the human position target.
Fig. 7 shows the result of human target recognition.
FIG. 8 shows simulation results of the error of the estimation of the target of the position of the human body.
Fig. 9 is a simulation result of the accuracy of human target recognition.
Detailed Description
The following detailed description of specific embodiments of the invention is provided, but it should be understood that the scope of the invention is not limited to the specific embodiments.
Throughout the specification and claims, unless explicitly stated otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element or component but not the exclusion of any other element or component.
As shown in fig. 1, the present invention mainly includes two aspects, i.e., person identification and person location. Person identification and person location can be divided into two phases, namely: an offline phase and an online phase. In the off-line stage, a binocular camera is used for shooting face image information of a person to be identified, and a (person target label, a face picture) database is established; and recording the position information of the shot point, extracting the depth information of the shot picture from the shot depth picture of the binocular camera, and establishing a (position information, depth picture information) database. And (4) carrying out offline classification learning on the database (the personnel target label and the human face picture) by using a convolutional neural network to obtain a personnel target identification classification model. And (4) training a (position information, image depth information) database by using a support vector machine, learning the corresponding relation between the position information and the depth information, and obtaining a regression model based on the position information. In the on-line stage, after the binocular camera shoots the pictures of the people, a face detection algorithm is used, the detected face images are processed, and the people are identified by utilizing a people target identification classification model. And processing the corresponding depth information picture, inputting the depth information picture into a regression model of the position information to obtain the specific position of the target from the camera, and further obtaining the target position.
As shown in fig. 2, in the invention, the face data is collected in the off-line stage, and the convolutional neural network is used to perform off-line classification learning on the (human target label, human face picture) database to obtain a human target identification classification model. In the on-line stage, after the PC detects the face data, the collected face data is preprocessed and then is transmitted to a trained personnel target recognition classification model to realize personnel recognition.
As shown in fig. 3, this is an off-line learning flow chart of the human target recognition classification model. The method comprises the steps of shooting face image information of a person to be recognized by using a binocular camera, establishing a (person target label, face picture) database, preprocessing the database, inputting the preprocessed database into a convolutional neural network for training, and obtaining a person target recognition classification model. The convolutional neural network part applies four convolutional layers, three pooling layers and one full-link layer.
As shown in fig. 4, in the invention, shooting point position information is recorded at an offline stage, depth information of a shot picture is extracted from a binocular camera shooting depth picture by using an image processing technology in OpenCV, and a (position information, depth picture information) database is established. And training the (position information and depth picture information) database by using a support vector machine, learning the corresponding relation between the position information and the depth picture information, and obtaining a regression model based on the position information. And in an online stage, acquiring video information by using a binocular camera at an unknown position, processing a corresponding depth information picture after people are detected, carrying out normalization processing on pixel information of the depth information picture, and obtaining a specific position of a target from the camera by using a regression model of position information so as to obtain the position of the target.
As shown in fig. 5, the present invention uses an image depth information picture photographed by a binocular camera.
As shown in fig. 6, the PC side human position target estimation result of the present invention. And displaying the distance between the target and the camera.
As shown in fig. 7, the PC side of the present invention shows the results of the person identification and the person position object estimation. And displaying the identification result of the target person and the distance between the target and the camera.
As shown in fig. 8, the average error of the distance estimation under the maximum number of training samples is not more than 5CM.
As shown in FIG. 9, the present invention achieves a recognition accuracy of 91% under the maximum number of training samples.
The foregoing description of specific exemplary embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable one skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications. It is intended that the scope of the invention be defined by the claims and their equivalents.
Claims (5)
1. A target identification and positioning method of a binocular camera is characterized by comprising the following steps: step 1: carrying out classification learning based on a Convolutional Neural Network (CNN) by using a face picture training database to obtain a classification model based on face recognition;
step 2: carrying out regression learning based on a Support Vector Machine (SVM) by utilizing the depth information of the picture to obtain a distance regression function, and constructing the relation between the depth information of the picture and the distance between a target and a camera;
and 3, step 3: shooting a target through a camera, and completing person target identification through a face classification model by using an obtained face picture;
and 4, step 4: meanwhile, the distance between the target and the camera is calculated by utilizing the depth information of the shot picture, so that personnel positioning is realized;
step 1: shooting face image information of a person to be identified by using a binocular camera, and establishing a person target label and a face image database; performing offline classification learning on the personnel target labels and the face image database by using a convolutional neural network to obtain a personnel target identification classification model;
step 2: recording the position information of a shooting point, extracting the depth information of a shooting picture from a binocular camera shooting depth picture by utilizing an image processing technology in OpenCV, and establishing a position information and depth picture information database; training a position information and depth picture information database by using a support vector machine, and learning the corresponding relation between the position information and the depth picture information to obtain a regression model based on the position information;
and step 3: acquiring motion information of a target by using a binocular camera, and after the binocular camera shoots a person picture, inputting a detected face image into the person target identification classification model in the step 1 by using a face detection algorithm through cutting processing to realize the identification of the person;
and 4, step 4: and (3) when the system detects the target, automatically processing the acquired depth picture information of the target, and then inputting the depth picture information into the regression model based on the position information in the step (2) to obtain the specific position of the target from the camera, further obtain the position of the target and realize the positioning of personnel.
2. The binocular camera target identifying and positioning method according to claim 1, wherein the steps 1 and 2 are an off-line stage, and the steps 3 and 4 are an on-line stage.
3. The binocular camera target recognition and positioning method of claim 1, wherein the face detection algorithm is a Cascade Classifier algorithm.
4. The binocular camera target identification and positioning method according to claim 3, wherein the offline stage and the online stage are specifically:
(1) The identification process of the personnel comprises the following steps:
the function of the module is as follows: firstly, calling a face detection module, cutting a detected face image after a face is detected, and then inputting the cut face image into a trained personnel target identification classification model, wherein the model can automatically identify the detected personnel to obtain an identification result; the convolutional neural network-based personnel target identification classification model is divided into two stages: an off-line stage and an on-line stage;
an off-line stage: shooting face image information of a person to be identified by using a binocular camera, and establishing a person target label and a face image database; performing offline classification learning on the personnel target labels and the human face image database by using a convolutional neural network to obtain a personnel target identification classification model;
an online stage: after the binocular camera shoots the pictures of the people, a human face detection algorithm is used, the detected human face images are cut, and the recognition of the people is realized by utilizing a human target recognition classification model;
(2) The positioning process of the personnel comprises the following steps:
the support vector machine-based personnel localization regression model is divided into two stages: an off-line phase and an on-line phase;
an off-line stage: recording the position information of a shooting point, extracting the depth information of a shooting picture from a binocular camera shooting depth picture, and establishing a position information and depth picture information database; training the position information and the depth picture information database by using a support vector machine, and learning the corresponding relation between the position information and the depth picture information to obtain a regression model based on the position information;
an online stage: the method comprises the steps of collecting video information at an unknown position by using a binocular camera, processing a corresponding depth information picture after people are detected, carrying out normalization processing on pixel information of the depth information picture, obtaining a specific position of a target from the camera by using a regression model of position information, and further obtaining the position of the target.
5. Personnel discernment and positioning system based on binocular camera, including personnel identification system and personnel positioning system, its characterized in that: the hardware equipment mainly comprises: the system comprises image acquisition equipment, algorithm processing equipment and display equipment;
an image acquisition device: use the binocular camera of the small size to accomplish image acquisition work, the model is S series, and the camera parameter is: a replaceable Standard M12 lens and a USB3.0 interface;
an algorithm processing device: the collected image data needs to be identified and positioned by personnel, the simulation experiment is realized on a computer, and the specific configuration is as follows: intel Core i5-7300HQ CPU 2.50GHz, internal memory 7.89G;
a system display device: the result after algorithm processing needs to be displayed and output, and the function is completed by adopting a display of a computer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910625272.3A CN110458025B (en) | 2019-07-11 | 2019-07-11 | Target identification and positioning method based on binocular camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910625272.3A CN110458025B (en) | 2019-07-11 | 2019-07-11 | Target identification and positioning method based on binocular camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110458025A CN110458025A (en) | 2019-11-15 |
| CN110458025B true CN110458025B (en) | 2022-10-14 |
Family
ID=68482685
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910625272.3A Active CN110458025B (en) | 2019-07-11 | 2019-07-11 | Target identification and positioning method based on binocular camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110458025B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111476126B (en) * | 2020-03-27 | 2024-02-23 | 海信集团有限公司 | Indoor positioning method, system and intelligent device |
| CN111462227A (en) * | 2020-03-27 | 2020-07-28 | 海信集团有限公司 | Indoor personnel positioning device and method |
| CN111461092B (en) * | 2020-06-19 | 2020-10-02 | 支付宝(杭州)信息技术有限公司 | Method, device and equipment for brushing face, measuring temperature and checking body |
| CN112164111B (en) * | 2020-09-10 | 2022-09-06 | 南京邮电大学 | Indoor positioning method based on image similarity and BPNN regression learning |
| CN112153736B (en) * | 2020-09-14 | 2022-07-26 | 南京邮电大学 | Personnel action identification and position estimation method based on channel state information |
| CN112184705B (en) * | 2020-10-28 | 2022-07-05 | 成都智数医联科技有限公司 | Human body acupuncture point identification, positioning and application system based on computer vision technology |
| CN113705376B (en) * | 2021-08-11 | 2024-02-06 | 中国科学院信息工程研究所 | Personnel positioning method and system based on RFID and camera |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066942A (en) * | 2017-03-03 | 2017-08-18 | 上海斐讯数据通信技术有限公司 | A kind of living body faces recognition methods and system |
| CN107145546A (en) * | 2017-04-26 | 2017-09-08 | 北京环境特性研究所 | Monitor video personnel's fuzzy retrieval method based on deep learning |
| CN107844744A (en) * | 2017-10-09 | 2018-03-27 | 平安科技（深圳）有限公司 | With reference to the face identification method, device and storage medium of depth information |
| CN108038455A (en) * | 2017-12-19 | 2018-05-15 | 中国科学院自动化研究所 | Bionic machine peacock image-recognizing method based on deep learning |
-
2019
- 2019-07-11 CN CN201910625272.3A patent/CN110458025B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107066942A (en) * | 2017-03-03 | 2017-08-18 | 上海斐讯数据通信技术有限公司 | A kind of living body faces recognition methods and system |
| CN107145546A (en) * | 2017-04-26 | 2017-09-08 | 北京环境特性研究所 | Monitor video personnel's fuzzy retrieval method based on deep learning |
| CN107844744A (en) * | 2017-10-09 | 2018-03-27 | 平安科技（深圳）有限公司 | With reference to the face identification method, device and storage medium of depth information |
| CN108038455A (en) * | 2017-12-19 | 2018-05-15 | 中国科学院自动化研究所 | Bionic machine peacock image-recognizing method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110458025A (en) | 2019-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110458025B (en) | Target identification and positioning method based on binocular camera | |
| CN110135249B (en) | Human behavior identification method based on time attention mechanism and LSTM (least Square TM) | |
| CN109887040B (en) | Moving target active sensing method and system for video monitoring | |
| WO2022188379A1 (en) | Artificial intelligence system and method serving electric power robot | |
| CN110969644B (en) | Personnel track tracking method, device and system | |
| CN105989369B (en) | Pedestrian re-identification method based on metric learning | |
| CN110751097B (en) | Semi-supervised three-dimensional point cloud gesture key point detection method | |
| CN104821010A (en) | Binocular-vision-based real-time extraction method and system for three-dimensional hand information | |
| CN105022999A (en) | Man code company real-time acquisition system | |
| CN104616002A (en) | Facial recognition equipment used for judging age groups | |
| US11361534B2 (en) | Method for glass detection in real scenes | |
| CN110443849B (en) | Target positioning method for double-current convolution neural network regression learning based on depth image | |
| CN109035307B (en) | Set area target tracking method and system based on natural light binocular vision | |
| CN109409250A (en) | A kind of across the video camera pedestrian of no overlap ken recognition methods again based on deep learning | |
| CN115082815B (en) | Tea bud picking point positioning method and device based on machine vision and picking system | |
| CN111460884A (en) | Multi-face recognition method based on human body tracking | |
| CN112207821B (en) | Target searching method of visual robot and robot | |
| CN114279433A (en) | Map data automatic production method, related device and computer program product | |
| CN115188066A (en) | Moving target detection system and method based on cooperative attention and multi-scale fusion | |
| CN108399630B (en) | Method for quickly measuring distance of target in region of interest in complex scene | |
| CN115376034A (en) | Motion video acquisition and editing method and device based on human body three-dimensional posture space-time correlation action recognition | |
| CN112767452B (en) | Active sensing method and system for camera | |
| CN114882072A (en) | High-speed correlation filtering target tracking method based on multi-channel image feature fusion | |
| CN114494857A (en) | Indoor target object identification and distance measurement method based on machine vision | |
| CN111399634A (en) | Gesture-guided object recognition method and device |
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 |