CN101437124A - Method for processing dynamic gesture identification signal facing (to)television set control - Google Patents

Method for processing dynamic gesture identification signal facing (to)television set control Download PDF

Info

Publication number
CN101437124A
CN101437124A CN 200810243904 CN200810243904A CN101437124A CN 101437124 A CN101437124 A CN 101437124A CN 200810243904 CN200810243904 CN 200810243904 CN 200810243904 A CN200810243904 A CN 200810243904A CN 101437124 A CN101437124 A CN 101437124A
Authority
CN
China
Prior art keywords
television
image
step
method
motion
Prior art date
Application number
CN 200810243904
Other languages
Chinese (zh)
Inventor
李春林
进 王
Original Assignee
三星电子(中国)研发中心;三星电子株式会社
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 三星电子(中国)研发中心;三星电子株式会社 filed Critical 三星电子(中国)研发中心;三星电子株式会社
Priority to CN 200810243904 priority Critical patent/CN101437124A/en
Publication of CN101437124A publication Critical patent/CN101437124A/en

Links

Abstract

The invention provides a method for processing a signal by dynamic gesture identification for television control, which relates to a signal processing method which detects motion in an image photographed by a television embedded camera, identifies actions of a hand and converts the action to a television control signal (such as channel switching, volume regulation, more complex menu operation and the like). The method comprises the following steps: acquiring a television user action image by the television built-in camera, analyzing the acquired image to acquire a motion region; extracting information of a hand target from the motion region to generate a track; and determining a command corresponding to the track to generate the television control command. The invention provides a method for controlling the television by a dynamic gesture through positioning the hand without manual assistance, which is to acquire a gesture language through a mode of detecting the motion and constructing a multiple objective body-models, and generate the corresponding television control command. The method provides a more improved mode of remote control for the field of gesture language television. The mode of remote control can significantly promote the development of the television application field.

Description

面向电视控制的动态手势识别信号处理方法 The gesture recognition for dynamic signal processing method of controlling the television

技术领域 FIELD

本发明涉及到从电视机内嵌摄像头摄取的图像中检测运动,对其中人手的动作进行识别并转换为电视控制信号(如频道切换、音量调整及更为复杂的菜单操作等) 的一种信号处理方法。 The present invention relates to A signal from the television camera embedded moving image pickup is detected, wherein the hand of the operation of the television recognized and converted into a control signal (e.g., channel switching, volume adjustment and a more complicated menu operation or the like) of Approach. 属于电视控制装置的技术领域。 TECHNICAL FIELD The television control belongs.

背景技术 Background technique

当前对电视进行远程遥控的模式主要有: Currently have the TV remote control modes are:

1. 红外电视远程遥控器(Infrared TV Remote Controller)遥控, 1. Infrared TV Remote Control (Infrared TV Remote Controller) remote control,

2. 类似鼠标的遥控器输入, 2. The mouse-like remote control input,

3. 采用手写板的遥控器手写输入,、 3. Using the remote control tablet ,, handwriting input

尽管已经有关于手语动作远程遥控电视的专利,如中国专利02101991. 6,但该专利基于对人体指尖红外信息的识别,从可行性,成本以及精确性上考虑都不是一个理想的技术方案。 Although there have been patents on sign language TV remote control operation, such as China patent 02101991.6, but the patent based on the recognition of the human fingertip infrared information, from the viewpoint of feasibility, cost and accuracy is not an ideal technical solution. 其他的类似专利也有一些关键的技术问题需要解决,如需要提高抗噪能力,或者需要人工辅助对手进行定位,等等。 Other similar patents are also some key technical problems to be solved, such as the need to improve noise immunity, or need artificial opponents positioning, and so on.

现有技术问题或要改善的地方 The prior art problems or to improve the place

1. 目前的电视遥控器无论从外形还是输入操作上都是千篇一律,已经让用户产生审美疲劳,不能突出产品在外形上的特色。 1. Current TV remote control in terms of appearance or input operations are stereotyped, already allows users to produce fatigued, can not highlight the product features in shape.

2. 随着电视功能越来越复杂,遥控器操作太过繁琐,消费者使用的功能还是基于80%-20%原则,即常用的功能仅仅是20%。 2. With the TV more complex functions, remote control operation is too cumbersome, consumers use the function is based on the principle of 80% -20%, that is commonly used function is only 20%.

3. 看电视本身是休闲娱乐,在需要对电视进行控制的时候(比如调节音量), 就需要手忙脚乱的找遥控器,而往往家里会有多个遥控器存在,电视,DVD, 空调,到底哪一个才是电视遥控器? 3. watch TV itself is entertainment, when you need to have the TV control (such as adjusting the volume), you need to rush to find a remote control, but often the family will have multiple remotes exist, TV, DVD, air-conditioning, which in the end is a TV remote control?

4. 当前手语电视的缺点:需要用户手动定位,不够人性化和智能化 4. The current shortcomings of sign language television: the need for users to manually locate, not humane and intelligent

发明内容技术问题:本发明的目的是实现一种结合运动检测和动态手势识别的电视控制方法,利用该方法可以通过操作者的手部动作(例如,挥手)来控制电视,比如调节音量,切换频道等,从而省去了寻找遥控器,操作遥控器的过程,将用户从对遥控器的依赖中解放出来,挥挥手即可实现对电视的控制。 Disclosure of Invention Technical Problem: The purpose of the present invention is to achieve a method for controlling television motion detection and dynamic binding gesture recognition, the method may use an operation by the operator's hand (e.g., wave) to control the TV, such as adjusting the volume, switching channel, etc., thereby eliminating the need to find a remote controller, remote control operation process, relieves the user from the remote controller in dependence out, waved to achieve control of the TV.

技术方案:本发明面向电视控制的动态手势识别信号处理方法涉及的是结合运动检测和手形识别的信号处理方法,该方法利用电视机内置摄像头采集电视用户动作图像,通过对采集到的图像进行分析得到运动区域,从运动区域中提取手目标的信息并生成轨迹,继而判断轨迹所对应的指令产生电视控制命令,该方法具体包含以下步骤- Technical Solution: Dynamic gesture recognition signal processing method for a television control of the present invention is a combination of motion detection and a signal processing method for hand shape recognition, the method utilizes the TV built-in camera TV capture user action image, analyze the collected images through deriving a motion region is extracted from the motion information of the target region of the hand and generates trajectories, and then determines the corresponding track to generate television control instruction command, the method specifically comprises the following steps -

步骤一:利用电视机中的内置摄像头采集视频数据,加入图像序列, Step 1: Use the built-in camera to capture video data of a television set, the image sequence is added,

步骤二:如果参考背景图像未构建,则用参考背景构建方法对歩骤一中采集到的图像序列构造参考背景图像, Step 2: If the reference background image is not constructed, the construction method of a step in the ho acquired image configuration with reference background reference background image sequence,

步骤三:如果参考背景图像已构建,则继续实时采集图像,将其与参考背景图像逐点作差,生成运动描述图像, Step 3: If the reference background image has been constructed, then continue to collect real-time image, which is the reference point by point as the difference between the background image, generating a motion description image,

步骤四:如果参考背景图像已构建,则开始实时采集图像,将其与参考背景图像比较,生成运动描述图像,并转化为二值运动掩模图像, Step four: If the reference background image has been constructed, the real-time image acquisition is started, which is compared with a reference background image, an image generating motion descriptors, and moving into a binary mask image,

步骤五:利用参考背景构建方法更新参考背景图像, Step five: Method using the reference background construct the reference background image,

步骤六:对二值运动图像进行目标分割和判定,利用多目标跟踪的方法构建和更新基于时空的多目标人体模型,其中每一个目标对应被分割出来的人体运动部位, Step Six: binary moving image object segmentation and determination, building and updating based on space-time multi-objective mannequin using multi-target tracking, in which each target portion corresponding to the body movement is divided out,

步骤七:如果多目标人体模型中存在已被跟踪的手目标,则更新该目标的轨迹后,检査轨迹是否与预设轨迹匹配并产生操作指令;否则检查是否有特定目标的轨迹符合需跟踪的手对象特征,如有,则将其标识, Step 7: If there is a hand that has been the target of multi-target tracking human body model, the update of the target track, checking the trajectory match and generate an operation command with a preset trajectory; otherwise check for traces of specific objectives in line with the need to track the object feature hand, if, it is identified,

步骤八:如产生操作指令,则将其送入电视内置软件模块,由其处理后完成相应操作, Step eight: The operation command is generated, it is sent to the TV built-in software module, after processing by the corresponding operation,

步骤九:重复步骤三,直到电视机被关闭。 Step 9: Repeat step three until the TV is turned off. 其中: among them:

步骤二中的参考背景构建方法是根据运动速度和幅度的不同选择不同的背景构建方法:当运动幅度和速度较大的情况下采用最新的一幅图像作为参考背景,运动平缓的情况下采用多帧图像平均的方法来计算参考背景。 Step two of the method of constructing the reference background construction method is selected according to different velocity and amplitude of the background: if the amplitude of motion is large and the velocity of a latest background image as a reference, the multi-movement and subdued average method to calculate the reference background frame image.

步骤四中二值运动掩模图像的生成方法是通过设定的门限值对运动描述图像进行二值化,将大于特定门限的点集合作为运动区域;其他点集合作为非运动区域。 Fourth step method for generating a binary mask image is moving by the threshold setting, the moving image binarization description, will be greater than a certain threshold point set as a motion area; other points as a set of non-motion area. 步骤六中利用多目标跟踪的方法为运动检测结果构建多目标人体模型,并且从 Sixth step of using a multi-target tracking method for constructing multi-objective model of human motion detection result, and from

中选择符合特定运动轨迹的手目标作为标识对象。 Select the hand target to meet specific trajectory as identify the object.

步骤七中对标识出的手对象运动轨迹采用基于轨迹库的模式匹配方法进行判 Step VII of the identified player object motion trajectory determination using pattern matching method for track-based library

别,与预设的操作指令轨迹进行比较,如有符合,产生相应的控制命令。 Respectively, with a preset trajectory comparison operation instruction, if met, generates a corresponding control command.

有益效果:本发明最大的贡献是提出了一种无需人工辅助对手进行定位的动态 Beneficial effects: the greatest contribution of this invention is to propose a rival without manual assistance dynamic positioning

手势控制电视的方法,该方法通过运动检测和构建多目标人体模型的方式获得用户的手语,并产生相应的电视控制指令。 The method of controlling the TV gesture, the motion detecting method and the building multi-objective mannequin manner to obtain the user's sign language, and generate a corresponding television control commands. 本发明为手语电视领域提供了更完善的远程遥控方式。 The present invention provides a better way remote control for the sign language television. 这种遥控方式将会显著地推动电视应用领域的发展。 This way the remote control will significantly promote the development of TV applications.

附图说明 BRIEF DESCRIPTION

图l示例的是手势识别信号处理流图,其中的关键步骤分别以数字①至数字⑧标识,与权利要求l中的步骤一至步骤八一一对应,如标识①为采集视频数据,标识② 为构建参考背景图像,标识③为生成运动描述图像,标识④为生成改进的运动掩模图像,t示识⑤为更新参考背景,标识⑥为生成多目标人体模型,标识⑦为手对象的轨迹判别,标识⑧为手势的识别及电视控制命令的生成。 Figure l is an example of a signal processing flow of FIG gesture recognition, wherein each of the digital key steps ① to ⑧ identification number, and the step claimed in claim l in a one-to eight steps, such as identification of captured video data ①, ② is identified Construction of the reference background image to generate a motion descriptors identifying image ③, ④ identify improved motion to generate the mask image, t is shown to update the reference background knowledge ⑤, ⑥ identifier to generate a multi-objective mannequin hand is identified ⑦ trajectory of the object is determined identifying ⑧ control command is generated gesture recognition and television.

具体实施方式' 一、 运动检测 DETAILED DESCRIPTION 'a motion detector

运动检测包含三个步骤: Detecting motion comprising three steps:

1. 参考背景构建 BACKGROUND 1. Reference Construction

在运动检测之前,首先要根据摄像单元拍摄到的图像进行参考背景构建,这里针对不同的情况提供两种背景构建的方法: Before the motion detection, the first to be constructed in accordance with the reference background image captured by the imaging unit, there is provided a method of constructing both contexts for different situations:

方法一:多帧图像统计法。 Method One: the multi-frame image statistics. 这种方法适用于手运动较慢的情况,具体实施方法是取多帧连续灰度图像,将图像中各点在时间上的平均值作为参考背景在该点的值。 This method is applicable to a case where the slower moving the hand, the specific method is to take a multi-frame embodiment of continuous tone images, the average image of each point in time as a reference value of the background points. 方法二:帧间参考法。 Method two: inter-reference method. 这种方法适用于手运动较快的情况,具体的实施方法是将上一次摄取的图像作为当前摄取图像的参考背景。 This method is suitable for fast hand movements, the particular embodiment of the previous method is taken as a reference background image of the current captured image.

2. 运动区域提取 2. The motion region extraction

参考背景构建成功后,通过将当前摄取图像中各点的灰度值与参考背景各点的、灰度值作差,以求出运动的轮廓。 After successful construction of the reference background, the current taken by the image gray value of each point with reference to the background point gray value for a difference in order of the contour motion. 具体的实施方法是-如果某一点的差值大于某个门限,则认为该点属于运动区域。 Specific embodiments is - if the difference is greater than a certain point threshold, it is considered that the point is in a motion area.

对于方法一构建的参考背景,采用统计门限。 For background reference method for constructing, using statistical threshold. 即"'=1 ,其中《为 I.e. '' = 1, where "as

i点的判决门限,A^)为第t帧用于构建参考背景的图像在i点的灰度值,A为参考背景在i点的灰度值。 Decision threshold point i, A ^) t-th frame is used to construct the reference background gray value image at the point i, A background reference gray value of point i.

对于方法二构建的背景,采用动态门限。 The method of constructing the background of the two, using the dynamic threshold.

通常当前摄取图像与参考背景的差值分布满足双峰特性,若yl和y 2分别是静止背景部分和运动区域的帧间差均值,则设定一个初始门限值,然后动态调节门限如下: Typically the current pickup image and the reference background difference bimodal distribution characteristics satisfied, and if yl y 2 differential frames are stationary background portion and the moving average of the region, an initial threshold is set, and dynamically adjusting the threshold as follows:

若当前门限G大于(ul+u2)/2,则将G减小一个步长;反之将G增加一个步长, 同时规定GE (Gmin, Gmax)。 If the current is greater than the threshold G (ul + u2) / 2, will be decreased by one step G; G conversely will increase by one step, while a predetermined GE (Gmin, Gmax).

3. 腐蚀运算 3. erosion operation

由2中得到的运动区域和非运动区域都需要进行腐蚀运算。 2 obtained by the motion area in the non-motion area and erosion operations need to be. 运动区域的腐蚀运算方法: Moving region etching operation method:

若某点属于运动对象,则当满足如下条件之一即可将该点同化为非运动区域: When the points belonging to a moving object, when one of the following conditions to the point of non-motion area assimilation:

① 与该点左右相邻的两个点属于非运动区域; ① point adjacent to the right and left two points belonging to non-motion area;

② 与该点上下相邻的两个点属于非运动区域; ② point adjacent to the upper and lower two points belonging to non-motion area;

③ 与该点左上和右下相邻的两个点属于非运动区域; ③ point adjacent to the upper left and lower right points belonging to two non-motion area;

④ 与该点左下和右上相邻的两个点属于非运动区域; 非运动区域的腐蚀运算方法: ④ two points adjacent to the lower left and upper-right point is non-motion area; etching method of calculating the non-motion area:

若某点属于非运动区域,则当满足如下条件之一即可将该点同化为非运动区域- If the point is in a non-motion area, one of the following conditions may be satisfied when the non-motion area point assimilation -

① 与该点左右相邻的两个点中至少一个点属于非运动区域; ① point adjacent to the right and left two points at least one point belonging to the non-motion area;

② 与该点上下相邻的两个点中至少一个点属于非运动区域; (D与该点左上和右下相邻的两个点中至少一个点属于非运动区域; ④与该点左下和右上相邻的两个点中至少一个点属于非运动区域; ② point adjacent to the upper and lower points in at least one point belonging to the non-motion area; (D with the upper left and lower right point of the two adjacent points in at least one point belonging to the non-motion area; ④ the lower left point, and right neighboring points in at least two points belonging to a non-motion area;

4. 背景更新 4. Background Update

在采用l中的方法一构建参考背景的情况下,为了保证参考背景的自适应性,需要对参考背景进行更新,从而能够发现忽然产生运动的物体,同时将停止运动的物体融入到参考背景中,2中提到的腐蚀运算可以保证这一点的实现。 In the case of a method of constructing l reference background, in order to ensure adaptive reference background, it is necessary to update the reference background, it is possible to generate sudden movement of objects found, while stopping the movement of objects into the background reference , 2 erosion operation mentioned can ensure this is achieved. 具体的更新方法为: Specific update methods are:

<formula>formula see original document page 8</formula>0 <Formula> formula see original document page 8 </ formula> 0

其中,Ii(t)为t时刻图像i点的灰度值;D^O表示i点属于非运动区域,DFl表示i点 Wherein, Ii (t) at time t is the gradation value of the image point i; D ^ O represents the point i belonging to non-motion area, DFl point represents i

属于运动区域;"是一个遗忘因子,参考值为0.8。其它定义与l中相同。 二、 动态手势识别 Belonging to the moving region; "is a forgetting factor, the reference value of 0.8 l the same as defined in the two other dynamic gesture recognition.

动态手势的识别包括如下三个部分:多目标跟踪,人体建模、轨迹判别和动态手势识别。 Identification of dynamic gestures include the following three parts: multi-target tracking, human modeling, a track is determined and the dynamic gesture recognition.

为了使系统具有更高的鲁棒性和识别能力,我们利用多目标跟踪的方法为运动检测结果构建人体模型,并在该模型的基础上进行轨迹判别和动态手势识别,同时还利用判定和识别的结果作为反馈对人体模型进行修正。 To make the system more robust and recognition, we use a method for multi-target tracking motion detection result mannequin construct, and determining the trajectory and dynamic gesture recognition on the basis of the model, while using determination and identification as a result of feedback on the human body model to be amended. 通常采用的多目标跟踪方 Commonly used in multi-target tracking party

法有联合概率数据关联滤波器(JPDAF)、多假设跟踪(MHT)算法、动态多维分配算法、无极卡尔曼滤波以及粒子滤波等,这些方法各有优劣,适合在不同场景和不同环境中使用,其原理在此均不作赘述。 Method has joint probabilistic data association filter (JPDAF), multiple hypothesis tracking (the MHT) algorithm, dynamic multi-dimensional allocation algorithm, Kalman filter and a particle filter Promise, these methods have advantages and disadvantages, suitable for use in different environments and different scenes , which is in principle not to repeat them here.

对实时视频数据的运动检测可以在每一帧中得到了一系列的运动对象,我们使用多目标跟踪的方法,为这些运动对象找到各自所对应的目标,并根据它们的轨迹为符合人体部位特征的有效目标建立人体模型。 Motion detection for real-time video data can be obtained in each frame of a series of moving objects, we use multi-target tracking, find the target corresponding to each of these moving objects, and in accordance with their trajectories to meet the characteristics of human body parts the goal of establishing an effective human model. 在多目标的人体模型中,每一个有效目标都对应有自己的时空轨迹,无论是在时间轴或是在空间坐标系中,有效目标都具有连续性的特点,且不同目标分别具有各自不同的表现形式。 Mannequin multi-target, each valid object corresponds to its own space-time trajectory, either in space or coordinate system, a valid object has continuity in the time axis characteristic, each having a respective different targets and different Manifestations. 例如手目标表现出来的运动性相对头目标更为强烈,而手目标与头目标之间的空间距离始终保持在某一区间;头目标的时空特征是常常会维持在同一高度而不常出现频繁的抖动,等等。 For example demonstrated by the target hand movement of the head relative to the target is more intense, and the spatial distance between the head and the goal is always to hand target maintained at a certain interval; spatial and temporal characteristics are often the first target will remain at the same level and not often appear frequently jitter, and so on. 这些特征为我们构建多目标人体模型提供了依据。 These features provide a basis for us to build multi-objective mannequin.

在多目标人体模型中,我们选取符合特定轨迹特征的对象作为被标识的手目标, 并对其轨迹进行识别,以判断是否产生控制指令。 In multi-objective manikin, we select the object that meet certain feature of the locus of the target is identified as the hand, and the identification of its trajectory, to determine whether a control instruction. 常用的轨迹识别方法有基于特征的识别、基于规则的识别以及基于规则库的模式匹配等,在轨迹判别和动态手势识别的过程中,我们可以利用预先设定的轨迹库作为被比较对象, 一旦目标轨迹符合预定轨迹库中特定轨迹的特征,则认为它们是同一轨迹,即可产生相应指令。 Commonly used methods are based on the trajectory recognition feature recognition, pattern recognition and rule-based matching rule-based library, and the like, and the dynamic process is determined in the track gesture recognition, we can use a preset library as the comparison target track, once wherein the predetermined target locus in line with the trajectory of the repository at a particular locus, they are considered the same track, to generate the corresponding instructions. 此外,轨迹库还可支持动态更新的功能,可按照用户的需要,录制特定功能的轨迹,以实现个性化的需求。 In addition, the track library also supports dynamic update functions, in accordance with the needs of users, track record specific functions to achieve individual needs.

在轨迹判别和动态手势识别结束后,判别和识别的结果还可以作为反馈被用于多目标人体模型的修正,当发生错误标识或者误判时,轨迹判别和识别的结果可以 After termination of the gesture and determining dynamic identification, identification and recognition results can also be used as the feedback correction multiobjective mannequin in the track, when an error occurs or false identification, track identification and recognition results

作为高权值的信息对已有标识进行校正,人体模型系统也在不断的修正中进一步完善。 Information as high weights to correct existing identity, human model system has also been revised further improvement.

Claims (5)

1. 一种面向电视控制的动态手势识别信号处理方法,其特征是:利用电视机内置摄像头采集电视用户动作图像,通过对采集到的图像进行分析得到运动区域,从运动区域中提取手目标的信息并生成轨迹,继而判断轨迹所对应的指令产生电视控制命令,该方法具体包含以下步骤:步骤一:利用电视机中的内置摄像头采集视频数据,加入图像序列,步骤二:如果参考背景图像未构建,则用参考背景构建方法对步骤一中采集到的图像序列构造参考背景图像,步骤三:如果参考背景图像已构建,则继续实时采集图像,将其与参考背景图像逐点作差,生成运动描述图像,步骤四:如果参考背景图像已构建,则开始实时采集图像,将其与参考背景图像比较,生成运动描述图像,并转化为二值运动掩模图像,步骤五:利用参考背景构建方法更新参考背景图像,步 1. A dynamic signal processing method of the gesture recognition control for the television, characterized in that: the use of a television TV built-in camera capture operation of the image the user, through the collected image is analyzed to obtain a motion area extracted from hand movement target area information and generate the track, and then determines the trajectory corresponding command generating television control commands, the method specifically includes the following steps: step 1: use the built-in camera in the television video data collection, added image sequence, step 2: If the reference background image is not Construction, the construction method of step one captured image sequence to construct the reference background image for the reference background, step 3: If the reference background image is constructed, continues real-time image acquisition, which is the reference background image point by point as the difference, generated described moving image step four: If the reference background image has been constructed, the real-time image acquisition is started, which is the reference background image, and generates moving image is described, and moving into a binary mask image, step five: constructed using the reference background The method of updating the reference background image, step 六:对二值运动图像进行目标分割和判定,利用多目标跟踪的方法构建和更新基于时空的多目标人体模型,其中每一个目标对应被分割出来的人体运动部位,步骤七:如果多目标人体模型中存在已被跟踪的手目标,则更新该目标的轨迹后,检查轨迹是否与预设轨迹匹配并产生操作指令;否则检查是否有特定目标的轨迹符合需跟踪的手对象特征,如有,则将其标识,步骤八:如产生操作指令,则将其送入电视内置软件模块,由其处理后完成相应操作,步骤九:重复步骤三,直到电视机被关闭。 Six: binary moving image object segmentation and determination, building and updating based on space-time multi-objective human model using multi-target tracking, in which each target corresponding to split out of the body motion component, Step 7: If multiple target body the presence of a chiral target model has been tracked, the target track is updated, check whether the locus matches the predetermined track and to generate an operation instruction; otherwise, check whether there is an object feature hand trajectory meet needs specific target tracking, and if so, then it is identified, step eight: the operation command is generated, it is sent to the TV built-in software module, by appropriate operation process is completed, step 9: step three is repeated, until the television set is turned off.
2. 根据权利要求1所述的一种面向电视控制的动态手势识别信号处理方法,其特征是步骤二中的参考背景构建方法是根据运动速度和幅度的不同选择不同的背景构建方法:当运动幅度和速度较大的情况下采用最新的一幅图像作为参考背景,运动平缓的情况下采用多帧图像平均的方法来计算参考背景。 According to one of the claim 1, the gesture recognition for dynamically controlling television signal processing method, wherein the step two of the reference background construction method is constructed according to different method of selecting a motion speed and magnitude of the background: when the motion using the latest reference background picture as a case of a large amplitude and speed, multi average reference background frame image using the calculated motion and subdued.
3. 根据权利要求1所述的一种面向电视控制的动态手势识别信号处理方法,其特征是歩骤四中二值运动掩模图像的生成方法是通过设定的门限值对运动描述图像进行二值化,将大于特定门限的点集合作为运动区域;其他点集合作为非运动区域。 According to one of the claim 1, the gesture recognition for Dynamic control TV signal processing method, wherein the generating step method the Fourth ho binary mask image by moving the threshold setting, the moving image description binarized, will be greater than a certain threshold set point as a motion area; other points as a set of non-motion area.
4.根据权利要求l所述的一种面向电视控制的动态手势识别信号处理方法,其特征是步骤六中利用多目标跟踪的方法为运动检测结果构建多目标人体模型,并且从中选择符合特定运动轨迹的手目标作为标识对象。 4. According to one of the claims l gesture recognition for Dynamic control TV signal processing method, wherein the step of utilizing six multi-target tracking method for constructing the multi-objective model of human motion detection result, and choose a particular motion compliance the trajectory of the hand as a target to identify the object.
5.根据权利要求1所述的一种面向电视控制的动态手势识别信号处理方法,其特征是步骤七中对标识出的手对象运动轨迹采用基于轨迹库的模式匹配方法进行判别,与预设的操作指令轨迹进行比较,如有符合,产生相应的控制命令。 According to one of the claim 1, the gesture recognition for dynamically controlling television signal processing method, wherein the step of identifying the seven target trajectory of the hand determination using pattern matching method for track-based library, and a preset comparing the command trajectory operation, if met, generates a corresponding control command.
CN 200810243904 2008-12-17 2008-12-17 Method for processing dynamic gesture identification signal facing (to)television set control CN101437124A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200810243904 CN101437124A (en) 2008-12-17 2008-12-17 Method for processing dynamic gesture identification signal facing (to)television set control

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200810243904 CN101437124A (en) 2008-12-17 2008-12-17 Method for processing dynamic gesture identification signal facing (to)television set control

Publications (1)

Publication Number Publication Date
CN101437124A true CN101437124A (en) 2009-05-20

Family

ID=40711319

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200810243904 CN101437124A (en) 2008-12-17 2008-12-17 Method for processing dynamic gesture identification signal facing (to)television set control

Country Status (1)

Country Link
CN (1) CN101437124A (en)

Cited By (55)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101853568A (en) * 2010-04-13 2010-10-06 鸿富锦精密工业(深圳)有限公司;鸿海精密工业股份有限公司 Gesture remote control device
CN101950222A (en) * 2010-09-28 2011-01-19 深圳市同洲电子股份有限公司 Handwriting input method, device and system for digital television receiving terminal
CN101951474A (en) * 2010-10-12 2011-01-19 冠捷显示科技(厦门)有限公司 Television technology based on gesture control
CN102012740A (en) * 2010-11-15 2011-04-13 中国科学院深圳先进技术研究院 Man-machine interaction method and system
CN102055925A (en) * 2009-11-06 2011-05-11 康佳集团股份有限公司 Television supporting gesture remote control and using method thereof
CN102073878A (en) * 2010-11-15 2011-05-25 上海大学 Non-wearable finger pointing gesture visual identification method
CN102136070A (en) * 2010-01-22 2011-07-27 财团法人工业技术研究院 Posture recognizing method and system as well as computer program product
CN102143312A (en) * 2011-03-24 2011-08-03 华为终端有限公司 Control method, device and system for object display
CN101719015B (en) 2009-11-03 2011-08-31 上海大学 Method for positioning finger tips of directed gestures
CN102193621A (en) * 2010-03-17 2011-09-21 三星电子(中国)研发中心 Vision-based interactive electronic equipment control system and control method thereof
CN102193626A (en) * 2010-03-15 2011-09-21 欧姆龙株式会社 Gesture recognition apparatus, method for controlling gesture recognition apparatus, and control program
CN102200834A (en) * 2011-05-26 2011-09-28 华南理工大学 television control-oriented finger-mouse interaction method
CN102221885A (en) * 2011-06-15 2011-10-19 青岛海信电器股份有限公司 Television, and control method and device thereof
CN102256099A (en) * 2011-06-20 2011-11-23 中兴通讯股份有限公司 Parameter control method and device
CN101777250B (en) 2010-01-25 2012-01-25 中国科学技术大学 General remote control device and method for household appliances
CN102375681A (en) * 2010-08-17 2012-03-14 Lg电子株式会社 The display apparatus and a control method
CN102457688A (en) * 2011-12-30 2012-05-16 四川长虹电器股份有限公司 Intelligent regulation method for volume and channel of television
CN102520786A (en) * 2011-09-13 2012-06-27 深圳市比维视创科技有限公司 Method and system for utilizing user action to realize control for electronic equipment by intelligent terminal
WO2012085923A1 (en) * 2010-12-24 2012-06-28 Hewlett-Packard Development Company, L. P. Method and system for classification of moving objects and user authoring of new object classes
CN102654792A (en) * 2011-03-04 2012-09-05 青岛海信电器股份有限公司 Controlled electronic equipment and control method thereof as well as intelligent television
CN102681658A (en) * 2011-01-06 2012-09-19 三星电子株式会社 Display apparatus controlled by motion and motion control method thereof
CN102682589A (en) * 2012-01-09 2012-09-19 西安智意能电子科技有限公司 System for distant control of controlled device
CN102789176A (en) * 2012-07-04 2012-11-21 北京捷通华声语音技术有限公司 Control method and system for household appliance terminal
CN102819751A (en) * 2012-08-21 2012-12-12 长沙纳特微视网络科技有限公司 Man-machine interaction method and device based on action recognition
WO2013037082A1 (en) * 2011-09-12 2013-03-21 Intel Corporation Using gestures to capture multimedia clips
US8417026B2 (en) 2009-12-30 2013-04-09 Industrial Technology Research Institute Gesture recognition methods and systems
CN103092332A (en) * 2011-11-08 2013-05-08 苏州中茵泰格科技有限公司 Digital image interactive method and system of television
TWI396442B (en) * 2009-05-21 2013-05-11 Chunghwa Telecom Co Ltd Application of gesture to recognize the gesture label of the Internet TV platform
CN103118297A (en) * 2013-01-22 2013-05-22 广东星海数字家庭产业技术研究院有限公司 Fatigue-prevention digital television system based on motion identification
CN103186358A (en) * 2011-12-27 2013-07-03 和鑫光电股份有限公司 Handheld electronic device capable of wirelessly transmitting images, and multimedia system formed by same
CN103295028A (en) * 2013-05-21 2013-09-11 深圳Tcl新技术有限公司 Gesture operation control method, gesture operation control device and intelligent display terminal
CN103294197A (en) * 2013-05-22 2013-09-11 深圳Tcl新技术有限公司 Terminal and gesture-operation-based method for realizing remote control on terminal
CN103442177A (en) * 2013-08-30 2013-12-11 程治永 PTZ video camera control system and method based on gesture identification
CN103500335A (en) * 2013-09-09 2014-01-08 华南理工大学 Photo shooting and browsing method and photo shooting and browsing device based on gesture recognition
US8810641B2 (en) 2011-06-15 2014-08-19 Hisense Hiview Tech Co., Ltd. Television, control method and control device for the television
WO2014172815A1 (en) * 2013-04-24 2014-10-30 Li Min Gesture television remote controller
CN104184941A (en) * 2014-06-27 2014-12-03 海信集团有限公司 Monitoring system
CN104219587A (en) * 2014-08-20 2014-12-17 深圳智意能电子科技有限公司 Method and device used for controlling application
CN104333794A (en) * 2014-11-18 2015-02-04 电子科技大学 Channel selection method based on depth gestures
CN104408444A (en) * 2014-12-15 2015-03-11 北京国双科技有限公司 Human body action recognition method and device
CN104424649A (en) * 2013-08-21 2015-03-18 株式会社理光 Method and system for detecting moving object
CN104486679A (en) * 2011-08-05 2015-04-01 三星电子株式会社 Method of controlling electronic apparatus and electronic apparatus using the method
CN104506586A (en) * 2014-12-09 2015-04-08 北京奇虎科技有限公司 Intelligent earphone system capable of regulating volume by gesture and regulation method
CN105009033A (en) * 2013-03-15 2015-10-28 高通股份有限公司 Interactive inputs for background task
CN105094301A (en) * 2014-05-16 2015-11-25 中兴通讯股份有限公司 Control method and device, and electronic equipment
CN105224090A (en) * 2010-12-27 2016-01-06 日立麦克赛尔株式会社 Image display device
CN105357568A (en) * 2015-11-09 2016-02-24 江苏惠通集团有限责任公司 Broadcasting source switching control method and device
CN105391964A (en) * 2015-11-04 2016-03-09 广东欧珀移动通信有限公司 Video data processing method and apparatus
CN105844263A (en) * 2010-12-24 2016-08-10 佳能株式会社 Summary view of video objects sharing common attributes
US9513711B2 (en) 2011-01-06 2016-12-06 Samsung Electronics Co., Ltd. Electronic device controlled by a motion and controlling method thereof using different motions to activate voice versus motion recognition
US9733895B2 (en) 2011-08-05 2017-08-15 Samsung Electronics Co., Ltd. Method for controlling electronic apparatus based on voice recognition and motion recognition, and electronic apparatus applying the same
CN107077609A (en) * 2014-11-06 2017-08-18 高通股份有限公司 Non-parametric model for detecting spatially different temporal modes
CN107493495A (en) * 2017-08-14 2017-12-19 深圳市国华识别科技开发有限公司 Interaction locations determine method, system, storage medium and intelligent terminal
CN108108709A (en) * 2017-12-29 2018-06-01 纳恩博(北京)科技有限公司 A kind of recognition methods and device, computer storage media
WO2019080780A1 (en) * 2017-10-24 2019-05-02 京东方科技集团股份有限公司 Gesture detection method, gesture processing device, and computer readable storage medium

Cited By (74)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI396442B (en) * 2009-05-21 2013-05-11 Chunghwa Telecom Co Ltd Application of gesture to recognize the gesture label of the Internet TV platform
CN101719015B (en) 2009-11-03 2011-08-31 上海大学 Method for positioning finger tips of directed gestures
CN102055925A (en) * 2009-11-06 2011-05-11 康佳集团股份有限公司 Television supporting gesture remote control and using method thereof
US8417026B2 (en) 2009-12-30 2013-04-09 Industrial Technology Research Institute Gesture recognition methods and systems
CN102136070A (en) * 2010-01-22 2011-07-27 财团法人工业技术研究院 Posture recognizing method and system as well as computer program product
CN102136070B (en) * 2010-01-22 2013-10-30 财团法人工业技术研究院 Posture recognizing method and system as well as computer program product
CN101777250B (en) 2010-01-25 2012-01-25 中国科学技术大学 General remote control device and method for household appliances
CN102193626B (en) * 2010-03-15 2015-01-21 欧姆龙株式会社 Gesture recognition apparatus, and method for controlling gesture recognition apparatus
CN102193626A (en) * 2010-03-15 2011-09-21 欧姆龙株式会社 Gesture recognition apparatus, method for controlling gesture recognition apparatus, and control program
CN102193621A (en) * 2010-03-17 2011-09-21 三星电子(中国)研发中心 Vision-based interactive electronic equipment control system and control method thereof
CN101853568A (en) * 2010-04-13 2010-10-06 鸿富锦精密工业(深圳)有限公司;鸿海精密工业股份有限公司 Gesture remote control device
CN102375681A (en) * 2010-08-17 2012-03-14 Lg电子株式会社 The display apparatus and a control method
CN102375681B (en) * 2010-08-17 2016-05-11 Lg电子株式会社 Display device and control method thereof
CN101950222A (en) * 2010-09-28 2011-01-19 深圳市同洲电子股份有限公司 Handwriting input method, device and system for digital television receiving terminal
CN101951474A (en) * 2010-10-12 2011-01-19 冠捷显示科技(厦门)有限公司 Television technology based on gesture control
CN102073878B (en) 2010-11-15 2012-11-07 上海大学 Non-wearable finger pointing gesture visual identification method
CN102073878A (en) * 2010-11-15 2011-05-25 上海大学 Non-wearable finger pointing gesture visual identification method
CN102012740A (en) * 2010-11-15 2011-04-13 中国科学院深圳先进技术研究院 Man-machine interaction method and system
WO2012085923A1 (en) * 2010-12-24 2012-06-28 Hewlett-Packard Development Company, L. P. Method and system for classification of moving objects and user authoring of new object classes
CN105844263B (en) * 2010-12-24 2019-08-23 佳能株式会社 The schematic diagram of the video object of shared predicable
CN105844263A (en) * 2010-12-24 2016-08-10 佳能株式会社 Summary view of video objects sharing common attributes
CN105224090A (en) * 2010-12-27 2016-01-06 日立麦克赛尔株式会社 Image display device
CN102681658A (en) * 2011-01-06 2012-09-19 三星电子株式会社 Display apparatus controlled by motion and motion control method thereof
US9513711B2 (en) 2011-01-06 2016-12-06 Samsung Electronics Co., Ltd. Electronic device controlled by a motion and controlling method thereof using different motions to activate voice versus motion recognition
US9398243B2 (en) 2011-01-06 2016-07-19 Samsung Electronics Co., Ltd. Display apparatus controlled by motion and motion control method thereof
CN102681658B (en) * 2011-01-06 2016-03-09 三星电子株式会社 By display device and the method for controlling operation thereof of action control
CN102654792A (en) * 2011-03-04 2012-09-05 青岛海信电器股份有限公司 Controlled electronic equipment and control method thereof as well as intelligent television
WO2012126400A1 (en) * 2011-03-24 2012-09-27 华为终端有限公司 Material object display control method, device and system
CN102143312A (en) * 2011-03-24 2011-08-03 华为终端有限公司 Control method, device and system for object display
CN102200834B (en) 2011-05-26 2012-10-31 华南理工大学 Television control-oriented finger-mouse interaction method
CN102200834A (en) * 2011-05-26 2011-09-28 华南理工大学 television control-oriented finger-mouse interaction method
CN102221885A (en) * 2011-06-15 2011-10-19 青岛海信电器股份有限公司 Television, and control method and device thereof
US8810641B2 (en) 2011-06-15 2014-08-19 Hisense Hiview Tech Co., Ltd. Television, control method and control device for the television
CN102221885B (en) * 2011-06-15 2013-06-19 青岛海信电器股份有限公司 Television, and control method and device thereof
WO2012174931A1 (en) * 2011-06-20 2012-12-27 中兴通讯股份有限公司 Parameter control method and device
CN102256099A (en) * 2011-06-20 2011-11-23 中兴通讯股份有限公司 Parameter control method and device
CN104486679A (en) * 2011-08-05 2015-04-01 三星电子株式会社 Method of controlling electronic apparatus and electronic apparatus using the method
US9733895B2 (en) 2011-08-05 2017-08-15 Samsung Electronics Co., Ltd. Method for controlling electronic apparatus based on voice recognition and motion recognition, and electronic apparatus applying the same
CN103828379A (en) * 2011-09-12 2014-05-28 英特尔公司 Using gestures to capture multimedia clips
WO2013037082A1 (en) * 2011-09-12 2013-03-21 Intel Corporation Using gestures to capture multimedia clips
CN102520786A (en) * 2011-09-13 2012-06-27 深圳市比维视创科技有限公司 Method and system for utilizing user action to realize control for electronic equipment by intelligent terminal
CN103092332A (en) * 2011-11-08 2013-05-08 苏州中茵泰格科技有限公司 Digital image interactive method and system of television
CN103186358B (en) * 2011-12-27 2016-03-16 和鑫光电股份有限公司 Can the hand-held electronic equipment of Wireless transceiver picture and the multimedia system of composition thereof
CN103186358A (en) * 2011-12-27 2013-07-03 和鑫光电股份有限公司 Handheld electronic device capable of wirelessly transmitting images, and multimedia system formed by same
CN102457688A (en) * 2011-12-30 2012-05-16 四川长虹电器股份有限公司 Intelligent regulation method for volume and channel of television
CN102682589B (en) * 2012-01-09 2015-03-25 西安智意能电子科技有限公司 System for distant control of controlled device
CN102682589A (en) * 2012-01-09 2012-09-19 西安智意能电子科技有限公司 System for distant control of controlled device
CN102789176B (en) * 2012-07-04 2015-08-05 北京捷通华声语音技术有限公司 A kind of household electrical appliance terminal control method and system
CN102789176A (en) * 2012-07-04 2012-11-21 北京捷通华声语音技术有限公司 Control method and system for household appliance terminal
CN102819751A (en) * 2012-08-21 2012-12-12 长沙纳特微视网络科技有限公司 Man-machine interaction method and device based on action recognition
CN103118297A (en) * 2013-01-22 2013-05-22 广东星海数字家庭产业技术研究院有限公司 Fatigue-prevention digital television system based on motion identification
CN105009033A (en) * 2013-03-15 2015-10-28 高通股份有限公司 Interactive inputs for background task
WO2014172815A1 (en) * 2013-04-24 2014-10-30 Li Min Gesture television remote controller
CN103295028B (en) * 2013-05-21 2018-09-04 深圳Tcl新技术有限公司 gesture operation control method, device and intelligent display terminal
CN103295028A (en) * 2013-05-21 2013-09-11 深圳Tcl新技术有限公司 Gesture operation control method, gesture operation control device and intelligent display terminal
CN103294197A (en) * 2013-05-22 2013-09-11 深圳Tcl新技术有限公司 Terminal and gesture-operation-based method for realizing remote control on terminal
CN104424649A (en) * 2013-08-21 2015-03-18 株式会社理光 Method and system for detecting moving object
US9552514B2 (en) 2013-08-21 2017-01-24 Ricoh Company, Ltd. Moving object detection method and system
CN103442177A (en) * 2013-08-30 2013-12-11 程治永 PTZ video camera control system and method based on gesture identification
CN103500335A (en) * 2013-09-09 2014-01-08 华南理工大学 Photo shooting and browsing method and photo shooting and browsing device based on gesture recognition
CN105094301A (en) * 2014-05-16 2015-11-25 中兴通讯股份有限公司 Control method and device, and electronic equipment
CN104184941A (en) * 2014-06-27 2014-12-03 海信集团有限公司 Monitoring system
CN104219587A (en) * 2014-08-20 2014-12-17 深圳智意能电子科技有限公司 Method and device used for controlling application
CN107077609A (en) * 2014-11-06 2017-08-18 高通股份有限公司 Non-parametric model for detecting spatially different temporal modes
CN104333794A (en) * 2014-11-18 2015-02-04 电子科技大学 Channel selection method based on depth gestures
CN104506586A (en) * 2014-12-09 2015-04-08 北京奇虎科技有限公司 Intelligent earphone system capable of regulating volume by gesture and regulation method
CN104408444A (en) * 2014-12-15 2015-03-11 北京国双科技有限公司 Human body action recognition method and device
CN105391964A (en) * 2015-11-04 2016-03-09 广东欧珀移动通信有限公司 Video data processing method and apparatus
CN105357568A (en) * 2015-11-09 2016-02-24 江苏惠通集团有限责任公司 Broadcasting source switching control method and device
CN105357568B (en) * 2015-11-09 2019-01-18 江苏惠通集团有限责任公司 A kind of control method and device of broadcast source switching
CN107493495B (en) * 2017-08-14 2019-12-13 深圳市国华识别科技开发有限公司 Interactive position determining method, system, storage medium and intelligent terminal
CN107493495A (en) * 2017-08-14 2017-12-19 深圳市国华识别科技开发有限公司 Interaction locations determine method, system, storage medium and intelligent terminal
WO2019080780A1 (en) * 2017-10-24 2019-05-02 京东方科技集团股份有限公司 Gesture detection method, gesture processing device, and computer readable storage medium
CN108108709A (en) * 2017-12-29 2018-06-01 纳恩博(北京)科技有限公司 A kind of recognition methods and device, computer storage media

Similar Documents

Publication Publication Date Title
Chen et al. Human action recognition using star skeleton
US8296151B2 (en) Compound gesture-speech commands
US8971629B2 (en) User interface system based on pointing device
US6754373B1 (en) System and method for microphone activation using visual speech cues
US8094881B2 (en) Action recognition apparatus and method, moving-object recognition apparatus and method, device control apparatus and method, and program
Liu et al. Hand gesture recognition using depth data
US10242255B2 (en) Gesture recognition system using depth perceptive sensors
CN100583124C (en) Method and system for automatic detection and tracking of multiple individuals using multiple cues
US6686844B2 (en) Human interface system using a plurality of sensors
KR101092820B1 (en) Lipreading and Voice recognition combination multimodal interface system
US9400548B2 (en) Gesture personalization and profile roaming
US20020089412A1 (en) Control system with user authentication
JP3978938B2 (en) Image projection apparatus and method
US6900790B1 (en) Information input apparatus, information input method, and recording medium
US20030156756A1 (en) Gesture recognition system using depth perceptive sensors
EP2367140B1 (en) Gesture recognition apparatus, method for controlling gesture recognition apparatus, and control program
US7308112B2 (en) Sign based human-machine interaction
JP2012518236A (en) Method and system for gesture recognition
US6360003B1 (en) Image processing apparatus
US9141196B2 (en) Robust and efficient learning object tracker
TWI411935B (en) System and method for generating control instruction by identifying user posture captured by image pickup device
US9278287B2 (en) Visual based identity tracking
CN101697199B (en) Detection method of head-face gesture and disabled assisting system using same to manipulate computer
US20120226981A1 (en) Controlling electronic devices in a multimedia system through a natural user interface
US8570371B2 (en) Object tracking apparatus, object tracking method, and control program

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)