CN106599812A - 3D dynamic gesture recognition method for smart home system - Google Patents

3D dynamic gesture recognition method for smart home system Download PDF

Info

Publication number
CN106599812A
CN106599812A CN201611104062.2A CN201611104062A CN106599812A CN 106599812 A CN106599812 A CN 106599812A CN 201611104062 A CN201611104062 A CN 201611104062A CN 106599812 A CN106599812 A CN 106599812A
Authority
CN
China
Prior art keywords
image
image processor
detection instrument
infrared
gesture
Prior art date
Application number
CN201611104062.2A
Other languages
Chinese (zh)
Inventor
史志晔
Original Assignee
苏州维盟韵联网络科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 苏州维盟韵联网络科技有限公司 filed Critical 苏州维盟韵联网络科技有限公司
Priority to CN201611104062.2A priority Critical patent/CN106599812A/en
Publication of CN106599812A publication Critical patent/CN106599812A/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/00335Recognising movements or behaviour, e.g. recognition of gestures, dynamic facial expressions; Lip-reading
    • G06K9/00355Recognition of hand or arm movements, e.g. recognition of deaf sign language
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/62Methods or arrangements for recognition using electronic means
    • G06K9/6201Matching; Proximity measures

Abstract

The invention relates to a 3D dynamic gesture recognition method for a smart home system. The method comprises the steps of establishing a platform in advance, wherein the platform includes a thermal imaging detector, a gesture image database and an image processor; performing the infrared image acquisition of a person who make gestures by the thermal imaging detector; uploading the data of a plurality of acquired infrared images to the image processor; pre-processing the plurality of uploaded infrared images by the image processor; detecting and segmenting human hands in the infrared images by the image processor; through the contour extraction of infrared images by the image processor, finding out palm areas; comparing the positions and the gestures of palms in the palm areas of the plurality of infrared images within a certain period of time by the image processor, and obtaining a gesture recognition result. The method is good in reliability, high in practicability, high in recognition efficiency, faster in processing speed, and quicker in system response. Based on the method, the requirements of real-time human-computer interaction can be met. The method is especially suitable for switching on or switching off household electrical appliances in the smart home system through gestures.

Description

A kind of 3D dynamic gesture identification methods for intelligent domestic system
Technical field
The present invention relates to a kind of gesture identification method, and in particular to a kind of 3D dynamic gestures for intelligent domestic system are known Other method.
Background technology
With the fast development of science and technology, studying the human-computer interaction technology HCI of natural harmony becomes Showed Very Brisk.Man-machine interaction From being progressively transferred to centered on computer, focus be put on man for technology, is the interaction technique of multimedia various modes, people's Interaction intention is naturally expressed as a multidimensional information vector, such as people in interacting with the external world while using gesture, language, and eye God etc..Can user produce well with virtual environment and interact, and a critically important research is exactly Gesture Recognition.
Computer vision technique has been working hard develop to the intelligent direction comparable to the wisdom of humanity, to be best understood from field Scape.Gesture Recognition has been used up for a long time 2D visions and is studied, but with the appearance of 3D sensor technologies, its application By increasingly extensive and variation.Because 2D visual scenes have in itself limitation, gesture recognition system must be various using other Information can just obtain the more preferable result comprising more useful information.When probability packet contains some motion trackings, depend merely on 2D also is difficult to obtain the result of relative gesture.
A kind of gesture identification method for being to be based on video, is the information pair arrived according to video acquisition in 3D gesture identification methods Gesture is analyzed, so as to realize the purpose of gesture identification.The method has equipment simple, the characteristics of cheap, but in handss Gesture identification need to process substantial amounts of data, and process is complicated, and software processes efficiency is low.In addition, the gesture identification method is also only used In a part of fields such as game, it is impossible to be effectively applied to as in the middle of mankind's service for life.
In consideration of it, it is that the present invention to be studied to provide a kind of 3D dynamic gesture identification methods for intelligent domestic system Problem.
The content of the invention
For the problems referred to above, it is an object of the invention to provide a kind of 3D dynamic hand gesture recognitions for intelligent domestic system Method, it is intended to improve 3D dynamic hand gesture recognition efficiency, and Gesture Recognition is applied in life staying idle at home.
To achieve these goals, the present invention is employed the following technical solutions:A kind of 3D for intelligent domestic system is dynamic Gesture identification method, one is built in advance is used to recognize the platform of gesture that the platform to include a thermal sensation imaging detection instrument, a gesture figure As data and an image processor, there are some specified images of gestures numbers for pre-depositing in the images of gestures data base According to the thermal sensation imaging detection instrument is connected by wired mode with described image processor, the 3D dynamic hand gesture recognition sides Method is operated according to the following steps:
The first step, infrared image acquisition is carried out using thermal sensation imaging detection instrument to the people for making gesture, when thermal sensation imaging detection instrument When capturing people and making action, multiple infrared images are automatically continuously gathered within a period of time, described a period of time can lead to Cross and preset in described image processor;
Multiple infrared picture datas for collecting are uploaded to described image processor by second step, the thermal sensation imaging detection instrument;
3rd step, described image sensor carries out pretreatment to the infrared image that thermal sensation imaging detection instrument is uploaded, logical by contrast The scope of ordinary person's body heat source is quickly found out the position of thermal source people in infrared image;
4th step, carries out carrying out the staff in infrared image by setting self adaptation dynamic depth threshold value in the image processor Detection and segmentation;
5th step, in the image processor, by the contours extract of image processor mid-infrared image palm area, and root is found Change according to a period of time, search out the region of each palm in described multiple infrared images;
6th step, in multiple infrared images of described image processor according to contrast within described a period of time in the region of palm The position of palm and gesture, and the specified gesture in advance having in the images of gestures after contrast and the images of gestures data base View data is compared, so as to obtain gesture identification result.
It is in the third step, further comprising the steps of in the pretreatment as the further improvement of such scheme:At image Reason device is filtered out and uploads the infrared image produced without gesture in data.
As the further improvement of such scheme, in the first step, when the thermal sensation imaging detection instrument does not capture people, Then automatically prolong shooting interval time of infrared image.
As the further improvement of such scheme, the 3D dynamic gesture identification methods also include a training step, with Before the platform of identification gesture comes into operation, specified gesture is stored in advance in images of gestures data base by the method trained View data.
Used as the further improvement of such scheme, the shooting direction of the thermal sensation imaging detection instrument can be with human body thermal source Direction carries out self angle adjustment.
As the further improvement of such scheme, in the first step, when thermal sensation imaging detection instrument captures people action is made When, the time interval of the thermal sensation imaging detection instrument crawl image is 30 milliseconds.
As the further improvement of such scheme, in the first step, when the thermal sensation imaging detection instrument does not capture people, The time interval of the thermal sensation imaging detection instrument crawl image is 60 milliseconds.
Used as the further improvement of such scheme, the platform also includes a photographic head.
Used as the further improvement of such scheme, the platform also includes a remote control and a remote control transceiver module, and this is distant Control transceiver module can receive the signal of remote control transmission, while remote control transceiver module can send a signal to remote control, so as to Information exchange is realized with remote control.
Operation principle of the present invention and effect are as follows:
The present invention relates to a kind of 3D dynamic gesture identification methods for intelligent domestic system, build in advance one for recognizing gesture Platform, the platform includes thermal sensation imaging detection instrument, images of gestures data base, image processor and makes the people of gesture, handss There are some specified images of gestures data for pre-depositing in gesture image data base, thermal sensation imaging detection instrument is by wired mode It is connected with image processor, is operated according to the following steps:The people for making gesture is carried out using thermal sensation imaging detection instrument infrared Image acquisition;Multiple infrared picture datas for collecting are uploaded to into image processor;Imageing sensor is to thermal sensation imaging detection The infrared image that instrument is uploaded carries out pretreatment;Carried out to infrared by setting self adaptation dynamic depth threshold value in the image processor Staff in image is detected and split;Palm area is found by the contours extract of image processor mid-infrared image;Figure As processor is according to the position of palm and gesture in the region of palm in multiple infrared images within a period of time of contrast, so as to Obtain gesture identification result.Good reliability of the present invention, practicality are high, recognition efficiency is high, and processing speed is very fast, and system is responded more Quickly, the requirement of real time human-machine interaction is disclosure satisfy that, is particularly well-suited in household be turned on and off household by gesture motion Electrical equipment.
Description of the drawings
Accompanying drawing described here is only used for task of explanation, and is not intended to by any way limit model disclosed in the present application Enclose.In addition, shape and proportional sizes of each part in figure etc. are only schematic, the understanding to the application is used to help, and It is not the shape and proportional sizes for specifically limiting each part of the application.Those skilled in the art, can under teachings of the present application To select various possible shapes and proportional sizes to implement the application as the case may be.In the accompanying drawings:
The step of accompanying drawing 1 is embodiment of the present invention schematic flow sheet.
Specific embodiment
Example below will be further illustrated the present invention.These embodiments are merely to illustrate the present invention, but not appointing Where formula limits the present invention.
Embodiment:A kind of 3D dynamic gesture identification methods for intelligent domestic system
Referring to accompanying drawing 1, one is built in advance is used to recognize the platform of gesture that the platform to include a thermal sensation imaging detection instrument, a gesture Image data base, an image processor and one make the people of gesture, have in the images of gestures data base and some deposit in advance The specified images of gestures data for entering, the thermal sensation imaging detection instrument presses following step by wired mode and described image processor Suddenly operated:
The first step, infrared image acquisition is carried out using thermal sensation imaging detection instrument to the people for making gesture, when thermal sensation imaging detection instrument When capturing people and making action, multiple infrared images are automatically continuously gathered within a period of time, described a period of time can lead to Cross and preset in described image processor;When thermal sensation imaging detection instrument captures people makes action, the thermal sensation imaging The time interval of survey meter crawl image is 30 milliseconds;When the thermal sensation imaging detection instrument does not capture people, the thermal sensation into As the time interval that survey meter captures image is 60 milliseconds.
Multiple infrared picture datas for collecting are uploaded to described image and are processed by second step, the thermal sensation imaging detection instrument Device.
3rd step, described image sensor carries out pretreatment to the infrared image that thermal sensation imaging detection instrument is uploaded, by right Than the position that the scope of usual human body thermal source is quickly found out thermal source people in infrared image, and filtered out by image processor Pass the infrared image produced without gesture in data.
4th step, is carried out to the staff in infrared image by setting self adaptation dynamic depth threshold value in the image processor Detected and split.
5th step, in the image processor, by the contours extract of image processor mid-infrared image palm area is found, And changed according to a period of time, search out the region of each palm in described multiple infrared images.
6th step, the area of palm in multiple infrared images of described image processor according to contrast within described a period of time The position of palm and gesture in domain, and by specifying for having in advance in the images of gestures after contrast and the images of gestures data base Images of gestures data are compared, so as to obtain gesture identification result.
In above step, also including a training step, before coming into operation for the platform for recognizing gesture, by training Method be stored in specified images of gestures data in images of gestures data base in advance.
Further, the platform also includes a remote control and a remote control transceiver module, and the remote control transceiver module can connect The signal of remote control transmission is received, while remote control transceiver module can send a signal to remote control, so as to realize information with remote control Interaction.
Further, the platform also includes a photographic head.
The present invention relates to a kind of 3D dynamic gesture identification methods for intelligent domestic system, build in advance a platform, should Platform includes thermal sensation imaging detection instrument, images of gestures data base and image processor, is operated according to the following steps:Using thermal sensation Imaging detection instrument carries out infrared image acquisition to the people for making gesture;Multiple infrared picture datas for collecting are uploaded to into image Processor;Imageing sensor carries out pretreatment to the infrared image for uploading;By in image processor to the people in infrared image Handss are detected and split;Palm area is found by the contours extract of image processor mid-infrared image;Image processor root Position and gesture according to palm in the region of palm in multiple infrared images of contrast within a period of time, knows so as to obtain gesture Other result.Good reliability of the present invention, practicality are high, recognition efficiency is high, and processing speed is very fast, and system response is more quick, can Meet the requirement of real time human-machine interaction, be particularly well-suited in household be turned on and off household electric appliances by gesture motion.
It should be noted that in the description of the present application, unless otherwise stated, " multiple " be meant that two or two with On.
The combination of element here, composition, part or step is described using term "comprising" or " including " it is also contemplated that The basic embodiment being made up of these elements, composition, part or step.Here by using term " can with ", it is intended to illustrate Described any attribute that " can with " includes all is optional.
Multiple element, composition, part or step can be provided by single integrated component, composition, part or step.Alternative Ground, single integrated component, composition, part or step can be divided into detached multiple element, composition, part or step.It is used for The open "a" or "an" of description element, composition, part or step is not said to exclude other elements, composition, part Or step.
It should be understood that above description is to illustrate rather than to be limited.By reading above-mentioned retouching State, many embodiments and many applications outside the example for being provided all will be for a person skilled in the art aobvious and easy See.Therefore, the scope of this teaching should not determine with reference to foregoing description, but should with reference to aforementioned claim and this The four corner of the equivalent that a little claim are possessed is determining.It is for comprehensive purpose, all articles and special with reference to including The disclosure of profit application and bulletin is all by reference to being incorporated herein.Theme disclosed herein is omitted in aforementioned claim Any aspect is not intended to abandon the body matter, also should not be considered as applicant the theme is not thought of as it is disclosed A part for application theme.
The a series of detailed description of those listed above is only for the feasibility embodiment of the application specifically Bright, they simultaneously are not used to limit the protection domain of the application, all equivalent implementations made without departing from the application skill spirit Or change should be included within the protection domain of the application.

Claims (9)

1. a kind of 3D dynamic gesture identification methods for intelligent domestic system, it is characterised in that:One is built in advance for recognizing The platform of gesture, the platform includes a thermal sensation imaging detection instrument, a gesture view data and an image processor, the gesture There are some specified images of gestures data for pre-depositing in image data base, the thermal sensation imaging detection instrument passes through wired side Formula is connected with described image processor, and the 3D dynamic gesture identification methods are operated according to the following steps:
The first step, infrared image acquisition is carried out using thermal sensation imaging detection instrument to the people for making gesture, when thermal sensation imaging detection instrument When capturing people and making action, multiple infrared images are automatically continuously gathered within a period of time, described a period of time can lead to Cross and preset in described image processor;
Multiple infrared picture datas for collecting are uploaded to described image processor by second step, the thermal sensation imaging detection instrument;
3rd step, described image sensor carries out pretreatment to the infrared image that thermal sensation imaging detection instrument is uploaded, logical by contrast The scope of ordinary person's body heat source is quickly found out the position of thermal source people in infrared image;
4th step, carries out carrying out the staff in infrared image by setting self adaptation dynamic depth threshold value in the image processor Detection and segmentation;
5th step, in the image processor, by the contours extract of image processor mid-infrared image palm area, and root is found Change according to a period of time, search out the region of each palm in described multiple infrared images;
6th step, in multiple infrared images of described image processor according to contrast within described a period of time in the region of palm The position of palm and gesture, and the specified gesture in advance having in the images of gestures after contrast and the images of gestures data base View data is compared, so as to obtain gesture identification result.
2. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that: It is further comprising the steps of in the pretreatment in three steps:Image processor is filtered out and uploaded in data without the infrared of gesture generation Image.
3. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that: In one step, when the thermal sensation imaging detection instrument does not capture people, then the shooting interval time of the infrared image that automatically prolongs.
4., according to the arbitrary described 3D dynamic gesture identification methods for intelligent domestic system of claim 1-3, its feature exists In:The 3D dynamic gesture identification methods also include a training step, before coming into operation for the platform for recognizing gesture, lead to The method for crossing training is stored in advance specified images of gestures data in images of gestures data base.
5. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that:It is described The shooting direction of thermal sensation imaging detection instrument can carry out self angle adjustment with the direction of human body thermal source.
6. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that: In one step, when thermal sensation imaging detection instrument captures people makes action, between the time of the thermal sensation imaging detection instrument crawl image It is divided into 30 milliseconds.
7. 3D dynamic gesture identification methods for intelligent domestic system according to claim 6, it is characterised in that: In one step, when the thermal sensation imaging detection instrument does not capture people, the thermal sensation imaging detection instrument captures the time interval of image For 60 milliseconds.
8. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that:It is described Platform also includes a photographic head.
9. 3D dynamic gesture identification methods for intelligent domestic system according to claim 1, it is characterised in that:It is described Platform also includes a remote control and a remote control transceiver module, and the remote control transceiver module can receive the signal of remote control transmission, Simultaneously remote control transceiver module can send a signal to remote control, so as to realize information exchange with remote control.
CN201611104062.2A 2016-12-05 2016-12-05 3D dynamic gesture recognition method for smart home system CN106599812A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611104062.2A CN106599812A (en) 2016-12-05 2016-12-05 3D dynamic gesture recognition method for smart home system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611104062.2A CN106599812A (en) 2016-12-05 2016-12-05 3D dynamic gesture recognition method for smart home system

Publications (1)

Publication Number Publication Date
CN106599812A true CN106599812A (en) 2017-04-26

Family

ID=58596944

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611104062.2A CN106599812A (en) 2016-12-05 2016-12-05 3D dynamic gesture recognition method for smart home system

Country Status (1)

Country Link
CN (1) CN106599812A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102402680A (en) * 2010-09-13 2012-04-04 株式会社理光 Hand and indication point positioning method and gesture confirming method in man-machine interactive system
US20130089237A1 (en) * 2011-10-10 2013-04-11 Invisage Technologies, Inc. Sensors and systems for the capture of scenes and events in space and time
CN103353935A (en) * 2013-07-19 2013-10-16 电子科技大学 3D dynamic gesture identification method for intelligent home system
CN104317391A (en) * 2014-09-24 2015-01-28 华中科技大学 Stereoscopic vision-based three-dimensional palm posture recognition interactive method and system
CN105589553A (en) * 2014-09-23 2016-05-18 上海影创信息科技有限公司 Gesture control method and system for intelligent equipment
JP2016148900A (en) * 2015-02-10 2016-08-18 嘉泰 小笠原 Electronic apparatus

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102402680A (en) * 2010-09-13 2012-04-04 株式会社理光 Hand and indication point positioning method and gesture confirming method in man-machine interactive system
US20130089237A1 (en) * 2011-10-10 2013-04-11 Invisage Technologies, Inc. Sensors and systems for the capture of scenes and events in space and time
CN103353935A (en) * 2013-07-19 2013-10-16 电子科技大学 3D dynamic gesture identification method for intelligent home system
CN105589553A (en) * 2014-09-23 2016-05-18 上海影创信息科技有限公司 Gesture control method and system for intelligent equipment
CN104317391A (en) * 2014-09-24 2015-01-28 华中科技大学 Stereoscopic vision-based three-dimensional palm posture recognition interactive method and system
JP2016148900A (en) * 2015-02-10 2016-08-18 嘉泰 小笠原 Electronic apparatus

Similar Documents

Publication Publication Date Title
JP6079832B2 (en) Human computer interaction system, hand-to-hand pointing point positioning method, and finger gesture determination method
Li et al. Joint distance maps based action recognition with convolutional neural networks
Lv et al. Multimodal hand and foot gesture interaction for handheld devices
Zhu et al. Evaluating spatiotemporal interest point features for depth-based action recognition
Zhang et al. EgoGesture: A new dataset and benchmark for egocentric hand gesture recognition
Zhou et al. A novel finger and hand pose estimation technique for real-time hand gesture recognition
Keskin et al. Real time hand pose estimation using depth sensors
Kumar et al. A multimodal framework for sensor based sign language recognition
Kamal et al. A hybrid feature extraction approach for human detection, tracking and activity recognition using depth sensors
CN106055091B (en) A kind of hand gestures estimation method based on depth information and correcting mode
Han et al. Enhanced computer vision with microsoft kinect sensor: A review
Jalal et al. Individual detection-tracking-recognition using depth activity images
Ramanathan et al. Human action recognition with video data: research and evaluation challenges
CN103294996B (en) A kind of 3D gesture identification method
Liu et al. Hand gesture recognition using depth data
Khan et al. Survey on gesture recognition for hand image postures
Hasan et al. Hand gesture modeling and recognition using geometric features: a review
Ren et al. Depth camera based hand gesture recognition and its applications in human-computer-interaction
Hackenberg et al. Lightweight palm and finger tracking for real-time 3D gesture control
Raheja et al. Real-time robotic hand control using hand gestures
Gu et al. Human gesture recognition through a kinect sensor
Kim et al. Simultaneous gesture segmentation and recognition based on forward spotting accumulative HMMs
CN103839040B (en) Gesture identification method and device based on depth image
Sarkar et al. Hand gesture recognition systems: a survey
Jain et al. Real-time upper-body human pose estimation using a depth camera

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: 20170426

RJ01 Rejection of invention patent application after publication