CN1769107A - Automobile anti-theft method based on human face identification technology - Google Patents

Automobile anti-theft method based on human face identification technology Download PDF

Info

Publication number
CN1769107A
CN1769107A CN 200410067778 CN200410067778A CN1769107A CN 1769107 A CN1769107 A CN 1769107A CN 200410067778 CN200410067778 CN 200410067778 CN 200410067778 A CN200410067778 A CN 200410067778A CN 1769107 A CN1769107 A CN 1769107A
Authority
CN
China
Prior art keywords
face
car
vehicle
recognition
driver
Prior art date
Application number
CN 200410067778
Other languages
Chinese (zh)
Other versions
CN100341732C (en
Inventor
欧阳合
林晓芸
史翠竹
史志存
周毅
罗霖
唐�谦
万凯
Original Assignee
上海杰得微电子有限公司
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 上海杰得微电子有限公司 filed Critical 上海杰得微电子有限公司
Priority to CN 200410067778 priority Critical patent/CN100341732C/en
Publication of CN1769107A publication Critical patent/CN1769107A/en
Application granted granted Critical
Publication of CN100341732C publication Critical patent/CN100341732C/en

Links

Abstract

The invention discloses a vehicle anti-thief method based on the face-identification technique, which comprises following steps: first collecting data; then combining the motion visual information to process face identification; at last, controlling the identified result. After the identification, the identified information is connected to the control bus of vehicle power system, while the identified result is illegal and the vehicle is not started, controlling the motor of vehicle to avoid the start, and while the vehicle is moving, the main control system sends the control signal to the speed control system to limit its mph and combine the vehicle-carrying GPS positioning system to send the alarm information of real-time position of vehicle, which can apply the tracing and catching on the thief.

Description

一种基于人脸认证技术的汽车防盗方法 Car alarm method based on face authentication technology

技术领域 FIELD

本发明涉及汽车电子领域中一种基于人脸认证技术的汽车防盗方法。 The present invention relates to the field of automotive electronics automobile anti-theft method based on human face authentication technology.

背景技术 Background technique

汽车电子作为动力学和电子学的结合体,已日益成为中国电子工业关注的热点之一,而作为安全保障的汽车防盗一直受到汽车制造商和社会保险业的重视。 Automotive electronics as a combination of dynamics and electronics, has increasingly become a focus of attention China's electronics industry, and as a car alarm security has been valued carmakers and social insurance industry. 随着汽车电子日新月异的发展,汽车防盗安全体系正在成为人们研究的焦点。 With the rapid development of automotive electronics, automobile anti-theft security system is becoming the focus of research. 目前汽车防盗技术正朝着自动化、智能化的方向发展。 Currently car anti-theft technology is moving automation, intelligent direction. 比较先进的汽车防盗技术包括电子钥匙式电子锁、生物锁(指纹锁和声控防盗锁)等。 More advanced vehicle anti-theft technology, including electronic key electronic locks, biometric locks (fingerprint lock and voice-activated anti-theft lock) and so on. 现有的防盗技术一般只在汽车启动时进行一次验证,未对驾车情况进行实时监控,对于汽车行驶过程中出现的劫车事件则显得无能为力。 Existing anti-theft technology is generally only when the vehicle is started once verified, is not real-time monitoring of the driving situation, for cars with carjacking incident occurred during the powerless.

而人脸识别技术由于具有主动性、非侵犯性、不涉及个人隐私、用户友好和方便等优点,是一种非常有竞争力的身份鉴别方法,在安全、保安和电子商务等方面有广泛应用和良好的产业化前景。 The face recognition technology because of the initiative, non-invasive and does not involve personal privacy, user friendliness and ease, is a very competitive authentication method, there is a wide range of applications in safety, security and e-commerce and good industrial prospects.

现有的人脸识别技术主要有:基于几何特征的方法、基于模板的方法和基于模型的方法。 Existing face recognition technology are: geometry feature based methods, template-based and model-based approach. 基于几何特征的方法是最早、最传统的方法,通常需要和其他算法结合才能有比较好的效果;基于模板的方法可以分为基于相关匹配的方法、特征脸方法、线性判别分析方法、奇异值分解方法、神经网络方法、动态连接匹配方法等。 The method is based on the geometric characteristics of the oldest, most traditional methods, usually in conjunction with other algorithms in order to have better results; template-based approach can be divided into correlation matching method based on feature face method, linear discriminant analysis, singular value decomposition method, neural network method, dynamic link matching method or the like. 基于模型的方法则有基于隐形马尔柯夫模型,主动形状模型和主动外观模型的方法等。 There is a method based on hidden Markov model, active shape model active appearance model and model-based approach. 现有人脸识别算法一般只是对视频中的静态图像进行识别,未充分利用视频中的运动信息,识别计算量大,效率有待提高。 Existing face recognition algorithms generally only still image video identify underutilized motion video information, identifying computing capacity, efficiency to be improved.

目前相关的专利有:(1)中国发明专利申请公开说明书CN1495658A(公开日:2004年5月12日)。 Currently related patents are: (1) Chinese invention patent document CN1495658A (Publication Date: May 12, 2004). 该专利申请侧重于驾驶员的安全问题,通过监视驾驶员的驾车行为和精神状态,如盗车、驾驶员疲劳和左顾右盼等,在验证出驾驶员身份不合法时只是给出远程通信呼叫报警而已,未对车辆进行控制、定位,不利于报警后对车辆的后续追踪。 This patent application focuses on the safety of the driver, by monitoring a driver's driving behavior and mental state, such as car theft, driver fatigue and the like looked around, while verifying the identity of the driver is not only legitimate remote communication call alert is given only not controlling the vehicle, positioning is not conducive to the follow-up alarm of the vehicle. 同时该专利中人脸识别部分的方法采用的是图像规范化处理、背景减除、KL变换求取特征脸等方法,与本发明的方法明显不同。 While the recognition process of this patent is used in the image portion of the normalization process, background subtraction, KL transform obtains Eigenface the like, and the method of the present invention is significantly different.

(2)中国发明专利申请公开说明书CN1414514A(公开日:2003年4月30日)。 (2) Chinese invention patent document CN1414514A (Publication Date: April 30, 2003). 该专利申请只是实现视频监视的功能,从应用场合及控制方法上,与本发明都不相同。 This patent application is to achieve the functions of the video monitor, control, and from the application methods, the present invention is not the same.

发明内容 SUMMARY

本发明要解决的技术问题是提供一种基于人脸认证技术的汽车防盗方法,利用提取运动信息的人脸识别对驾驶员身份进行验证,同时用验证的结果启动汽车主控系统,保证只有授权人员具有驾驶权利,达到对驾驶员的实时监控,提高汽车的防盗性能。 The present invention to solve the technical problem is to provide a vehicle anti-theft method of face authentication technology based on the use of face recognition to extract motion information to verify the identity of the driver, and start the car with the main control system to verify the results to ensure that only authorized who has the right to drive, to achieve real-time monitoring of the driver, the car's performance to improve security.

为解决上述技术问题,本发明的一种基于人脸认证技术的汽车防盗方法,包括如下步骤:首先进行数据采集,使用带红外功能的视频采集设备获取视频数据;然后,进行人脸识别,识别时充分利用了运动视频信息;最后,对识别结果进行控制,若检测到驾驶人员是合法驾驶人员,则赋予驾驶者驾车的权利,同时将合法的信号传送到汽车动力系统的控制总线,总线得到这一信号后,安装正常模式允许驾驶员对汽车进行各种操作;若检测的结果不是人脸图像,则应进一步判别是否是本车内景图像,若是车内景图像,则说明车内无人,发动机不能启动;若不是本车内景图像,则有可能是驾驶人员戴面具进行驾驶,此时若汽车未启动,则通过汽车控制总线控制发动机,不允许发动机启动;若汽车处于行驶过程中,则通过控制换档控制部件对汽车进行限速, To solve the above technical problem, the present invention is a method of automobile anti-theft face authentication technology based, comprises the following steps: data acquisition, using a video capture device with infrared function acquires the video data; then, face recognition, identification when the full advantage of the motion video information; Finally, the recognition result to control, upon detecting the driver is a legitimate driver, claimed the driver to drive is given, while transmitting valid signals to the vehicle power system control bus, the bus to give after this signal, the normal mode allows the driver to install the various operations of the car; if the result is not the face image detected should be further determined whether the present vehicle is a background image, if the background image inside the car, then the unmanned vehicle, the engine can not be started; if not this car background image, it is possible that the driver wearing masks were driving this time if the car does not start, through the car engine control bus, do not allow the engine is started; if the car is running, then car speed limit by controlling the shift control member, 时主控系统通过汽车总线启动车载无线报警装置,向车主和公安系统110报警中心发出报警信号,通过车载GPS定位系统给出实时汽车位置信息;若检测到驾驶人员是非法驾驶人员,系统提示进入下一阶段的辅助识别;辅助输入密码识别功能仅在汽车没有行驶的状况下有效,验证通过后则不用再次输入密码直到下次汽车停车为止,再次行驶必须再次输入密码或者被识别人为合法驾驶者;若人脸识别及辅助密码验证均未通过,立即向车主和公安系统110报警中心发出报警信号;通过视频采集设备实时采集驾驶员图像数据,每隔一段时间进行新一轮的人脸识别,这样就可以达到对车内驾驶员的实时监控。 When the master vehicle bus system is initiated by the vehicle-mounted wireless alarm device, an alarm signal to the alarm center 110 and the public security system owner, the vehicle position information is given in real time by the onboard GPS positioning system; if the driver is detected illegal driver, prompted to enter the assist in identifying the next stage; auxiliary input password identification function is only available in the car is not driving condition, after verification by then without entering the password again until the next date car parking, travel must enter the password again be recognized again or man-made legitimate driver ; if no recognition by password authentication and an auxiliary, an alarm signal to the alarm center 110 and the public security system owners immediately; real-time data by the image acquisition driver video capture device, a new one at intervals of face recognition, in this way we can achieve real-time monitoring of the vehicle driver.

本发明的有益效果是,将人脸识别用于汽车防盗这一特定应用场合,侧重于身份验证后与汽车主控系统的连接与控制。 Advantageous effect of the invention is that the car alarm recognition for this particular application, focusing on the car is connected to the main control system after authentication. 车辆的行驶状态直接从汽车的速度检测系统获取,有利于装置的成本控制。 Vehicle driving state directly from the vehicle speed detection system that is conducive to cost control device. 在身份验证后,将验证信息与汽车动力系统控制总线相连接。 After authentication, the authentication information with the vehicle power system is connected to a control bus. 当验证结果为非法时,若汽车未启动,则控制汽车发动机,不允许其启动。 When the verification result is illegal when, if the car does not start, automobile engine control, not allowed to start. 若汽车处于行驶状态,则主控系统发控制信号给速度控制系统,限制汽车时速,并结合车载GPS定位系统给出实时车辆位置的报警信息,以利于及时追踪并捕获盗车者。 If the car is traveling, the main control system sent a control signal to the speed control system, restrictions on vehicle speed, combined with an alarm system Vehicular GPS vehicle position information in real time, to facilitate tracking and timely capture car theft.

本发明的汽车防盗方法采用人脸识别技术,具有主动性、非侵犯性、不涉及个人隐私、用户友好和方便,而且更灵活、安全、可靠,达到实时防盗的目的。 Automobile anti-theft method of the invention uses face recognition technology, proactive, non-invasive and does not involve personal privacy, user-friendly and convenient, but also more flexible, safe, reliable, achieve real-time anti-theft purposes.

附图说明 BRIEF DESCRIPTION

下面结合附图和具体实施方式对本发明作进一步详细的说明:图1是本发明的方法流程图;图2是图1中进行人脸识别的预处理流程图;图3是图1中进行人脸识别的人脸检测算法流程图;具体实施方式本发明所述的一种基于人脸认证技术的汽车防盗方法,将人脸识别技术和汽车主控系统相结合用于汽车防盗系统中,在汽车车内安装人脸识别汽车防盗系统FRMGS(Face Recognition for Mobile GuardSystem),通过视频采集系统实时采集驾驶人员的视频图像,进行人脸检测和识别,识别的结果通过汽车总线传递给汽车主控系统,由主控系统控制相应部件工作,从而决定是否赋予驾驶的权利。 The present invention will be further described in detail in conjunction with accompanying drawings and specific embodiments: FIG. 1 is a flowchart of a method according to the present invention; FIG. 2 is a flowchart of recognition pre-processing performed in FIG. 1; FIG. 3 is a human in FIG. face detection algorithm is a flowchart of a face recognition; the one of the embodiment of the present invention, face authentication method for vehicle anti-theft technology based on the face recognition technology and combine the main control system for automotive vehicle anti-theft system, automotive interior mounting face recognition car alarm system FRMGS (face recognition for Mobile GuardSystem), capture video images in real time by the video driver acquisition system, face detection and recognition, the recognition result is transmitted to the main control system via the vehicle bus cars , the control by the main control system of the respective working member, to determine whether to give driving claimed.

如图1所示,本发明的方法包括如下步骤。 1, the method of the present invention comprises the following steps.

首先,进行数据采集,例如可以使用视频采集设备实时采集视频数据。 Firstly, data collection, for example, using real-time video capture device to capture video data. 该视频采集设备是带红外功能的摄像机或摄像头等采集设备,主要目的是当车内光线不足时,能自动开启红外功能,使拍摄的效果仍能满足人脸识别的要求。 The video capture device is a camera or cameras with infrared capabilities acquisition equipment, the main purpose of the car is when there is insufficient light, it can automatically turn on the infrared function, the effect of shooting still meet the requirements of recognition.

然后,对采集的视频数据进行人脸识别。 Then, the video data captured for face recognition. 人脸识别主要包含:预处理、人脸检测和人脸识别三部分。 Main recognition comprising: preprocessing, face detection and recognition of three parts.

预处理。 Pretreatment. 将人脸识别算法用于汽车防盗系统中,对系统的实时性的要求比较高,而且人在开车过程中,人脸图像不是一成不变的,而是随着角度和光线的变化而不断变化,所以对图像进行人脸识别前必须进行必要的预处理,使图像的亮度和尺寸符合识别的要求。 The face recognition algorithm used in automobile anti-theft systems, real-time system requirements are relatively high, and people driving the process, the face image is not static, but changes with the angle and light are constantly changing, so pretreatment must be necessary for image recognition, brightness and size of the image corresponding to the identification requirements. 本发明的方法中进行预处理时,采用的是光线补偿和尺寸归一化处理,同时也不做背景的减除操作。 When the preprocessing method of the present invention, light is used to compensate and normalize the size, nor do the subtraction operation while the background. 光线补偿方法可以采用图像直方图均衡化处理,尺寸归一化则是对图像大小进行缩放,使得拍摄的人脸图像大小与数据库中人脸大小一致。 Light compensation method may be employed histogram equalization process, the size of normalization is scaling the image size, so that a person's face image photographed faces in the size of the database the same size. 预处理过程如图2所示。 Pretreatment process shown in FIG.

人脸检测。 Face Detection. 如图3所示,本发明采用肤色模型的方法进行人脸检测。 As shown in FIG. 3, the present invention employs a method of skin color model for face detection. 肤色是人脸的重要信息,不依赖于面部的细节特征,对于旋转、表情等变化情况都能适用,具有相对的稳定性并且和大多数背景物体的颜色相区别。 Color is important information for human face does not depend on details of the features of the face, for the case where the rotation changes, facial expressions and the like can be applied, and the relative stability and color of the most distinguished from the background object. 本发明采用归一化的rb(红-蓝)色度空间、高斯肤色模型进行肤色检测。 The present invention employs normalized RB (Red - blue) color space, the Gaussian model for skin color detection.

首先定义归一化的rb色度空间,将三维的RGB(红绿蓝色度空间)降为二维。 First, the definition of normalized rb color space, the three-dimensional RGB (red, green and blue of space) is reduced to two dimensions. 在这个二维平面上,肤色的区域很集中,可以用高斯模型来描述该分布;其次,进行训练得到该分布的中心;然后,根据所考察的象素离该中心的远近得到其和肤色的相似度,形成一个原图的相似度分布图,对该分布图二值化,以最终确定肤色的区域。 In this two-dimensional plane, the color region is focused, can be described with this Gaussian distribution model; secondly, to obtain training center of the distribution; then, according to which the considered pixel and the color obtained from the distance of the center of similarity, forming a picture of the distribution of similarity, the profile binarized to determine the final skin color region.

在检测出肤色象素后,需要根据它们在色度上的相似性和空间上相关性分割出可能的区域,同时利用区域的几何特征或灰度特征进行人脸验证,以排除其他色彩类似肤色的物体。 After detecting the color pixels, face verification is required based on the geometric features of their relevance to the divided regions may be similarity in spatial and chroma, while the use of regional or gray feature, to the exclusion of other similar skin color It objects.

人脸识别。 Face Recognition. 进行人脸识别时,首先建立授权人员的人脸数据库,然后建立待检测人脸与数据库中人脸的对应关系,进行判别。 When face recognition, face first established a database of authorized personnel, and establish a corresponding relationship with the database to be detected human face human face, discriminate. 可以广泛采用各种成熟的人脸识别算法,如Eigenface(特征脸方法)、Fisherface(线性判别方法)、神经网络算法以及各种新的人脸识别算法等,以提高人脸识别率,而不把算法限制为KL变换一种。 It can be widely employed in a variety of sophisticated recognition algorithms, such as Eigenface (eigenfaces method), the Fisherface (linear discriminant method), a neural network algorithm and a variety of new face recognition algorithms, in order to improve the recognition rate without the KL transform algorithm is limited to one kind.

在充分利用现有人脸识别技术的基础上,本发明还在人脸识别过程中充分利用了视频中的运动信息。 Making full use of the prior art face recognition, face recognition process of the present invention also take advantage of the motion video information. 传统的人脸识别方法对于视频中每帧图像的处理是独立进行的,没有利用相邻帧的运动信息。 The traditional method of face recognition processing for the video image of each frame is carried out independently, there is no adjacent frames using the motion information. 本发明充分利用了相邻帧图像之间的运动信息。 The present invention makes full use of the motion information between adjacent image frames. 在检测到一张人脸后,通过判断视频流中相邻帧图像运动的大小,或者是否发生了场景切换来决定是否进行下一轮的识别,只有在大的运动的时候,才进行下一轮的人脸识别,这样可以提高识别的准确度,减少由于人脸的角度及表情变化引起的误判,同时也减少了计算量。 Upon detecting a human face, if a scene change occurs is determined by the size of the frame image motion video stream adjacent to, or to determine whether the next round of identification, only when the large motion, before the next the wheel face recognition, which can improve the accuracy of recognition, since the angle and reduce false facial expression changes caused while also reducing the amount of calculation.

最后,对识别结果进行控制。 Finally, control on the recognition result. 将识别结果与汽车主控系统进行连接与控制。 The recognition result with the vehicle master control connected to the control system. 例如:将人脸识别结果通过汽车数据总线传至主控系统,通过汽车动力系统的控制总线(如高速CAN总线、低速CAN总线、LIN总线等),连接车内的汽车电子控制装置和车载电子装置,达到对发动机、油门及车载报警装置的控制。 For example: The face recognition results transmitted vehicle data bus master system, vehicle power system via a control bus (CAN bus such as high speed, low speed CAN bus, the LIN bus, etc.), connected to the vehicle electronic control devices and automotive vehicle electronic means to control of the engine throttle and vehicle alarm device.

汽车电子包含两部分,一是汽车电子控制装置,如发动机、底盘、车身控制等;二是车载汽车电子装置,如汽车信息系统、导航系统、车载通信系统等。 Automotive comprising two parts, the automotive electronic control device, such as the engine, chassis, body control; Second automotive vehicle electronic device, such as a vehicle information system, a navigation system, in-vehicle communication systems. 通讯线将各种汽车电子装置连接成为一个网络,通过数据总线发送和接收信息。 Various communication lines automotive electronics connected as a network, information transmitted and received via the data bus. 通过数据总线可以访问任何一个电子控制装置。 You can access any electronic control device via a data bus. 本发明就是将上述两类汽车电子有机结合起来。 The present invention has the above-described two types of automotive electronics combine. 应用本发明的方法把人脸识别系统作为车载汽车电子装置,通过汽车控制总线,将人脸识别的结果与汽车电子控制装置进行信息的联系与沟通,从而达到对汽车各部件的控制。 The method of the present invention is applied as a face recognition system automotive vehicle electronics, vehicle control bus via the recognition result and the automobile electronic control device associated with the communication of information, so as to control the respective parts of the automobile.

若系统检测到驾驶人员是合法驾驶人员,则将控制权交给汽车动力系统的控制总线(如高速CAN总线、低速CAN总线、LIN总线等),总线得到这一信息后,控制汽车发动机启动,允许驾驶。 If the system detects that the driver is a legitimate driver, it will control to the vehicle power system control bus (such as high speed CAN bus, low-speed CAN bus, LIN bus, etc.), the bus get this information, control the car engine starts, allowed to drive. 同时反光镜、方向盘、油门、座椅位置和车载音响等都可按驾驶者的意愿任意调整。 Meanwhile mirrors, steering wheel, the accelerator, the car stereo seat position and so on can be arbitrarily adjust the driver's wishes.

若检测的结果不是人脸图像,则应进一步判别是否是本车内景图像,若是车内景图像,则说明车内无人,发动机不启动;若不是本车内景图像,则有可能是驾驶人员戴面具进行驾驶,此时若汽车未启动,则通过汽车控制总线控制发动机,不允许发动机启动;若汽车处于行驶过程中,则通过控制换档控制部件对汽车进行限速,同时主控系统通过汽车总线启动车载无线报警装置,向车主和公安系统110报警中心发出报警信号。 If the test result is not a facial image, it should be further determine whether this car background image, if the car scene image, then no one inside the car, the engine does not start; if not this car background image, it is possible that the driver wearing for driving the mask, if the car does not start at this time, the control bus through the car engine, the engine start is not allowed; if the car is traveling, the car speed is performed by controlling the shift control member, while the main control system through the car bus start vehicle wireless alarm device, an alarm signal to the alarm center 110 and the public security system owners.

若检测到驾驶人员是非法驾驶人员,如果此时汽车处于停止状态,系统提示进入下一阶段的辅助识别(例如,输入密码等)。 When the driver detects that the driver is illegal, if the vehicle is in a stopped state at this time, the system prompted to enter the next phase of the auxiliary identification (e.g., a password, etc.). 本发明将人脸识别验证与辅助验证相结合,可以在不改变人脸数据库的情况下,提高增加临时驾驶员时的灵活性。 The present invention is recognition and verification auxiliary authentication combined, can be changed without face database, the temporary flexibility is increased when the driver. 若密码验证通过,则赋予驾驶权利。 If the password is verified, it gives the driver the right. 否则,拒绝其驾驶,将相应的信息通过汽车动力系统的控制总线(如高速CAN总线、低速CAN总线、LIN总线等)传递给相应的设备,如汽车引擎及报警装置等。 Otherwise, reject the driver, the corresponding information (such as high speed CAN bus, CAN bus low, the LIN bus, etc.) to the corresponding device via the vehicle power system control bus, such as automobile engine and alarm devices. 若汽车未启动,则控制汽车发动机,不允许其启动;若汽车处于行驶状态,则通过主控系统限制汽车的时速,同时主控系统通过汽车总线启动车载无线报警装置,向车主和公安系统110报警中心发出报警信号。 If the car does not start, the automobile engine control, which is not allowed to start; if the vehicle is in a traveling state, the vehicle speed limit by the main control system, while the master system start-vehicle wireless alarm device via the vehicle bus, to vehicle owners and the public security system 110 the alarm center an alarm signal. 公安系统接到报警信号后,结合被盗车辆的车载GPS定位系统,可以及时捕获盗车者。 Public security system after receiving the alarm signal, combined with on-board GPS positioning system stolen vehicle, timely capture car theft. 本发明将人脸识别验证与辅助验证相结合,可以弥补识别系统的灵活性不足,比如汽车临时给其他人驾驶的情况下,可以仅仅简单设置密码并设定密码的有效期限等即可,不需要额外建立驾驶者面相数据。 The present invention will be insufficient recognition verification and validation auxiliary combination, can make identification system flexibility, such as a case where the temporary car driving others, may only be set simply set the password and password expiration date can not the need for additional data to establish the driver's face.

在本发明中,通过视频采集系统实时采集驾驶员图像数据,每隔一段时间进行新一轮的人脸识别(时间间隔可以根据实际应用需要动态修改),这样就可以达到对车内驾驶员的实时监控。 In the present invention, the image data acquired by the video capture driver system in real time, at intervals of a new recognition round (a time interval depending on the application can dynamically modify), so that the vehicle can reach on the driver real time monitoring.

Claims (7)

1.一种基于人脸认证技术的汽车防盗方法,包括如下步骤:首先进行数据采集,然后,进行人脸识别,最后,对识别结果进行控制,其特征在于:若检测到的驾驶人员合法,则赋予驾驶者驾车的权利,同时将合法的信号传送到汽车动力系统的控制总线,所述总线得到该信号后,安装正常模式允许驾驶员对汽车进行各种操作;若检测的结果不是人脸图像,则应进一步判别是否是本车内景图像,若是,则说明车内无人,发动机不能启动;若不是,则有可能是驾驶人员戴面具进行驾驶;此时,若汽车未启动,则通过汽车控制总线控制发动机,不允许发动机启动;若汽车处于行驶过程中,则通过控制换档控制部件对汽车进行限速,同时主控系统通过汽车控制总线启动车载无线报警装置,向车主和公安系统110报警中心发出报警信号,通过车载GPS定位系统给出 1. A vehicle anti-theft method based on human face authentication technology, comprising the following steps: First, data acquisition, and then, face recognition, finally, controls the recognition result, wherein: when the detected driver legitimate, after the rights conferred driver driving, while transmitting a signal to the legitimate automotive powertrain control bus, said bus signal obtained, a normal mode mounting allows the driver to perform various operations of the car; if the result of the detection face is not the image should be further determine whether this car background image, if, then no one inside the car, the engine can not be started; if not, it is possible that the driver wearing masks were driving; this time, if the car does not start, through control bus automobile engine, the engine start is not allowed; if the car is traveling, the car speed is performed by controlling the shift control means, start the vehicle while the host system through the car radio alarm control bus, to vehicle owners and the public security system alarm center an alarm signal 110, given by the onboard GPS positioning system 时汽车位置信息;若检测的驾驶人员非法,则提示进入辅助识别,在汽车没有行驶的状况下,辅助输入密码进行识别,验证通过后允许驾驶员对汽车进行各种操作,直到下次汽车停车为止,再次行驶必须再次输入密码或者被识别人为合法驾驶者;若人脸识别及辅助密码验证均未通过,立即向车主和公安系统110报警中心发出报警信号;通过视频采集设备实时采集驾驶员图像数据,每隔一段时间进行新一轮的人脸识别,实现对车内驾驶员的实时监控。 The car location information; if detected driving illegally, you are prompted to enter assist in identifying, in the car is not driving condition, auxiliary input password identification, after verification by allowing the driver to perform various operations on the car until the next car parking So far, again with a password must be entered again or be artificially identified legitimate driver; if no recognition by password authentication and an auxiliary, an alarm signal to the alarm center 110 and the public security system owners immediately; real-time image of the driver acquired by the video capture device data, every once in a while a new round of face recognition, real-time monitoring of the vehicle driver.
2.如权利要求1所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述数据采集,使用带红外功能的视频采集设备获取视频数据。 2. An automobile according to claim 1 theft prevention method based on the face authentication technique, wherein: said data acquisition, using a video capture device with infrared function acquires video data.
3.如权利要求1所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述人脸识别包括预处理、人脸检测和人脸识别三个步骤。 One of the claim 1 car alarm face authentication method based technology, wherein: said pre-treatment comprises a recognition, face detection and recognition three steps.
4.如权利要求3所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述预处理,采用光线补偿和尺寸归一化方法进行处理,同时不做背景的减除操作;光线补偿方法可以采用图像直方图均衡化处理。 4. An automobile according to claim 3 theft prevention method based on the face authentication technique, wherein: said pretreatment, using light and dimensions normalization compensation processing method, and not the background subtraction operation ; light compensation method may use an image histogram equalization.
5.如权利要求3所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述人脸检测,采用肤色模型的方法进行人脸检测,首先定义归一化的rb色度空间,将三维的RGB降为二维,在该二维平面上,用高斯模型来描述该红一蓝色度分布情况;其次,进行训练得到该分布的中心;然后,根据所考察的象素离该中心的远近得到其和肤色的相似度,形成一个原图的相似度分布图,再按照一定的规则对该分布图二值化,以最终确定肤色的区域;在检测出肤色象素后,需要根据其在色度上的相似性和空间上相关性分割出可能的区域,同时利用该区域的几何特征或灰度特征进行人脸验证,以排除色彩类似肤色的物体。 5. An automobile according to claim 3 theft prevention method based on the face authentication technique, wherein: the face detection, skin color model of the method for face detection, first define the normalized color rb space, an RGB three-dimensional reduced to a two-dimensional, in the two-dimensional plane, with a Gaussian model used to describe the distribution of a blueness red; secondly, to obtain training center of the distribution; then, according to the considered pixel FIG similarity distribution obtained from the distance of the center of their similarity and color, to form a picture, and then according to certain rules the value of the two profiles, in order to determine the final skin color region; after detecting the skin color pixel , the need to their similarity in color and spatial correlation may be divided region, while the use of geometric features or characteristics of the gray region for face authentication, similar to exclude skin color object.
6.如权利要求3所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述人脸识别,首先建立授权人员的人脸数据库,然后建立待检测人脸与数据库中人脸的对应关系,进行判别;人脸识别算法可以采用特征脸方法、线性判别方法或神经网络算法。 6. An automobile according to claim 3 theft prevention method based on the face authentication technique, wherein: said recognition, first create a database the face of an authorized person, and the establishment of a database to be detected human face the corresponding relationship of the face, discriminate; facial feature recognition algorithm may be employed a method, a linear discriminant method or a neural network algorithm.
7.如权利要求3或6所述的一种基于人脸认证技术的汽车防盗方法,其特征在于:所述人脸识别中充分利用相邻帧图像之间的运动信息,在检测到一张人脸后,通过判断视频流中相邻帧图像运动的大小,或者是否发生了场景切换来决定是否进行下一轮的识别,只有在检测到大幅度运动的时候,才进行下一轮的人脸识别,以提高识别的准确度,减少由于人脸的角度及表情变化引起的误判,同时也减少计算量。 7. An claim 3 or claim 6 vehicle anti-theft face authentication method based technology, wherein: said full face recognition using the motion information between adjacent image frames, a detection of after the face, if a scene change occurs is determined by the size of the frame image motion video stream adjacent to, or to determine whether the next round of identification only when significant motion is detected, it is carried out under a person face recognition, recognition to improve accuracy, reduce false positives due to the angle of the face and facial expression changes caused, while also reducing the amount of calculation.
CN 200410067778 2004-11-03 2004-11-03 Automobile anti-theft method based on human face identification technology CN100341732C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200410067778 CN100341732C (en) 2004-11-03 2004-11-03 Automobile anti-theft method based on human face identification technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200410067778 CN100341732C (en) 2004-11-03 2004-11-03 Automobile anti-theft method based on human face identification technology

Publications (2)

Publication Number Publication Date
CN1769107A true CN1769107A (en) 2006-05-10
CN100341732C CN100341732C (en) 2007-10-10

Family

ID=36750725

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200410067778 CN100341732C (en) 2004-11-03 2004-11-03 Automobile anti-theft method based on human face identification technology

Country Status (1)

Country Link
CN (1) CN100341732C (en)

Cited By (39)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101863229A (en) * 2010-06-04 2010-10-20 奇瑞汽车股份有限公司 Engine start switch
CN101890938A (en) * 2010-08-10 2010-11-24 北京中科三环高技术股份有限公司;南京大陆鸽高科技股份有限公司;盐城赛福汽车零部件有限公司 Electric vehicle anti-theft alarm system and using method thereof
CN101902619A (en) * 2010-06-22 2010-12-01 浙江天鸿汽车用品有限公司 Vehicle-mounted intelligent identity recognizing and monitoring system
CN101934779A (en) * 2010-08-31 2011-01-05 中山大学 Intelligent vehicle anti-theft method and system
CN102224510A (en) * 2008-11-21 2011-10-19 罗伯特·博世有限公司 Device for creating and/or processing an object signature, monitoring device, method and computer program
CN102358259A (en) * 2011-09-16 2012-02-22 兰州大学 Automobile anti-theft automatic control system based on biological information identification
CN102446249A (en) * 2010-10-15 2012-05-09 鸿富锦精密工业(深圳)有限公司 Electronic device and method for acquiring illegal user information of same
CN102529887A (en) * 2012-02-15 2012-07-04 奇瑞汽车股份有限公司 Automobile anti-burglary method and automobile anti-burglary device
CN102750526A (en) * 2012-06-25 2012-10-24 黑龙江科技学院 Identity verification and recognition method based on face image
CN102034334B (en) 2009-09-28 2012-12-19 财团法人车辆研究测试中心 Driver monitoring method and monitoring system thereof
CN101506020B (en) 2006-08-31 2013-01-02 株式会社东海理化电机制作所 Hybrid car travel mode setting device
CN102855675A (en) * 2012-09-07 2013-01-02 三一重工股份有限公司 Engineering machinery, and authentication method, controller and system thereof
CN102975669A (en) * 2012-11-23 2013-03-20 安徽师范大学 Driver identity identification device and method
CN103095771A (en) * 2011-11-04 2013-05-08 上海博泰悦臻网络技术服务有限公司 Vehicle monitor method, device and system and vehicle terminal
CN103180179A (en) * 2010-09-13 2013-06-26 胡夫电子有限公司 Sensor device for motor vehicle
CN103253234A (en) * 2013-05-23 2013-08-21 无锡伊佩克科技有限公司 Vehicle anti-theft positioning system
CN103324904A (en) * 2012-03-20 2013-09-25 凹凸电子(武汉)有限公司 Face recognition system and method thereof
CN103612617A (en) * 2013-12-03 2014-03-05 沈阳工业大学 Automobile anti-theft anti-robbery system
CN103738296A (en) * 2013-12-30 2014-04-23 深圳市广和通实业发展有限公司 Automobile safety management system and method
CN103793681A (en) * 2012-10-30 2014-05-14 原相科技股份有限公司 Apparatus and method for identifying and confirming user and vehicle central control system using the apparatus
CN103921759A (en) * 2014-04-18 2014-07-16 深圳市威富安防有限公司 Intelligent anti-theft vehicle starting method and system
CN104097611A (en) * 2014-07-30 2014-10-15 中山艺展装饰工程有限公司 Automobile anti-theft system based on human face identification technology multi-identity authentication
CN104268525A (en) * 2014-09-24 2015-01-07 厦门蓝斯通信股份有限公司 Vehicle monitoring method and system based on face recognition
CN104554146A (en) * 2014-12-24 2015-04-29 苏州福丰科技有限公司 Anti-theft electric vehicle based on face recognition system
CN104859590A (en) * 2015-05-20 2015-08-26 张彬彬 GPS car alarm
CN105320863A (en) * 2015-10-28 2016-02-10 宝鸡市公安局交通警察支队 Driver identity verification system and method
CN105427642A (en) * 2015-12-29 2016-03-23 徐承柬 Vehicle management and control system
CN105550666A (en) * 2016-01-22 2016-05-04 大连楼兰科技股份有限公司 Alarming system and method for preventing juveniles from driving cars
CN105711547A (en) * 2016-01-25 2016-06-29 北京小米移动软件有限公司 Automobile thievery preventing method and device and device for preventing automobile from being stolen
CN105930781A (en) * 2016-04-14 2016-09-07 北京天诚盛业科技有限公司 Driver authorization controller and authorization control method using controller
CN106347300A (en) * 2016-08-29 2017-01-25 乐视控股(北京)有限公司 Method and apparatus for starting vehicle based on biological recognition
CN106710145A (en) * 2016-12-29 2017-05-24 清华大学苏州汽车研究院(吴江) Guided driver tiredness prevention method
CN107323421A (en) * 2017-06-30 2017-11-07 合肥虎俊装饰工程有限公司 A kind of multi-purpose vehicle(MPV) intelligent safety and defence system
CN107539276A (en) * 2017-07-04 2018-01-05 芜湖市振华戎科智能科技有限公司 Car man carrier interconnected system based on CNN
CN108357462A (en) * 2018-02-09 2018-08-03 北京车和家信息技术有限公司 The method and vehicle of vehicle unlock
WO2018218559A1 (en) * 2017-05-31 2018-12-06 深圳市永恒丰科技有限公司 Fingerprint identification based bicycle safety management method and system
CN109059947A (en) * 2018-07-04 2018-12-21 深圳明创自控技术有限公司 A kind of high performance onboard navigation system
CN109693644A (en) * 2018-11-26 2019-04-30 北京建筑大学 A kind of VATS Vehicle Anti-Theft System and method with multi-biological characteristic authentication
CN110027509A (en) * 2019-04-10 2019-07-19 上海工程技术大学 A kind of vehicular theft-prevention monitoring device and method based on recognition of face

Cited By (44)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101506020B (en) 2006-08-31 2013-01-02 株式会社东海理化电机制作所 Hybrid car travel mode setting device
CN102224510A (en) * 2008-11-21 2011-10-19 罗伯特·博世有限公司 Device for creating and/or processing an object signature, monitoring device, method and computer program
CN102034334B (en) 2009-09-28 2012-12-19 财团法人车辆研究测试中心 Driver monitoring method and monitoring system thereof
CN101863229A (en) * 2010-06-04 2010-10-20 奇瑞汽车股份有限公司 Engine start switch
CN101902619A (en) * 2010-06-22 2010-12-01 浙江天鸿汽车用品有限公司 Vehicle-mounted intelligent identity recognizing and monitoring system
CN101890938A (en) * 2010-08-10 2010-11-24 北京中科三环高技术股份有限公司;南京大陆鸽高科技股份有限公司;盐城赛福汽车零部件有限公司 Electric vehicle anti-theft alarm system and using method thereof
CN101890938B (en) * 2010-08-10 2013-07-10 北京中科三环高技术股份有限公司 Electric vehicle anti-theft alarm system and using method thereof
CN101934779A (en) * 2010-08-31 2011-01-05 中山大学 Intelligent vehicle anti-theft method and system
CN103180179B (en) * 2010-09-13 2016-07-13 胡夫·许尔斯贝克和福斯特有限及两合公司 Sensor device for automobile
CN103180179A (en) * 2010-09-13 2013-06-26 胡夫电子有限公司 Sensor device for motor vehicle
CN102446249A (en) * 2010-10-15 2012-05-09 鸿富锦精密工业(深圳)有限公司 Electronic device and method for acquiring illegal user information of same
CN102358259A (en) * 2011-09-16 2012-02-22 兰州大学 Automobile anti-theft automatic control system based on biological information identification
CN103095771A (en) * 2011-11-04 2013-05-08 上海博泰悦臻网络技术服务有限公司 Vehicle monitor method, device and system and vehicle terminal
CN102529887A (en) * 2012-02-15 2012-07-04 奇瑞汽车股份有限公司 Automobile anti-burglary method and automobile anti-burglary device
CN103324904A (en) * 2012-03-20 2013-09-25 凹凸电子(武汉)有限公司 Face recognition system and method thereof
CN102750526A (en) * 2012-06-25 2012-10-24 黑龙江科技学院 Identity verification and recognition method based on face image
CN102855675A (en) * 2012-09-07 2013-01-02 三一重工股份有限公司 Engineering machinery, and authentication method, controller and system thereof
CN102855675B (en) * 2012-09-07 2015-06-10 三一重工股份有限公司 Engineering machinery, and authentication method, controller and system thereof
CN103793681A (en) * 2012-10-30 2014-05-14 原相科技股份有限公司 Apparatus and method for identifying and confirming user and vehicle central control system using the apparatus
CN103793681B (en) * 2012-10-30 2017-12-15 原相科技股份有限公司 User identifies and confirmed device, method and the automobile central control system using the device
CN102975669A (en) * 2012-11-23 2013-03-20 安徽师范大学 Driver identity identification device and method
CN102975669B (en) * 2012-11-23 2016-04-13 安徽师范大学 A kind of driver identity means of identification and recognition methods thereof
CN103253234A (en) * 2013-05-23 2013-08-21 无锡伊佩克科技有限公司 Vehicle anti-theft positioning system
CN103612617A (en) * 2013-12-03 2014-03-05 沈阳工业大学 Automobile anti-theft anti-robbery system
CN103738296A (en) * 2013-12-30 2014-04-23 深圳市广和通实业发展有限公司 Automobile safety management system and method
CN103921759A (en) * 2014-04-18 2014-07-16 深圳市威富安防有限公司 Intelligent anti-theft vehicle starting method and system
CN104097611A (en) * 2014-07-30 2014-10-15 中山艺展装饰工程有限公司 Automobile anti-theft system based on human face identification technology multi-identity authentication
CN104268525A (en) * 2014-09-24 2015-01-07 厦门蓝斯通信股份有限公司 Vehicle monitoring method and system based on face recognition
CN104554146A (en) * 2014-12-24 2015-04-29 苏州福丰科技有限公司 Anti-theft electric vehicle based on face recognition system
CN104859590A (en) * 2015-05-20 2015-08-26 张彬彬 GPS car alarm
CN105320863A (en) * 2015-10-28 2016-02-10 宝鸡市公安局交通警察支队 Driver identity verification system and method
CN105427642A (en) * 2015-12-29 2016-03-23 徐承柬 Vehicle management and control system
CN105550666A (en) * 2016-01-22 2016-05-04 大连楼兰科技股份有限公司 Alarming system and method for preventing juveniles from driving cars
CN105711547A (en) * 2016-01-25 2016-06-29 北京小米移动软件有限公司 Automobile thievery preventing method and device and device for preventing automobile from being stolen
CN105930781A (en) * 2016-04-14 2016-09-07 北京天诚盛业科技有限公司 Driver authorization controller and authorization control method using controller
CN106347300A (en) * 2016-08-29 2017-01-25 乐视控股(北京)有限公司 Method and apparatus for starting vehicle based on biological recognition
CN106710145A (en) * 2016-12-29 2017-05-24 清华大学苏州汽车研究院(吴江) Guided driver tiredness prevention method
WO2018218559A1 (en) * 2017-05-31 2018-12-06 深圳市永恒丰科技有限公司 Fingerprint identification based bicycle safety management method and system
CN107323421A (en) * 2017-06-30 2017-11-07 合肥虎俊装饰工程有限公司 A kind of multi-purpose vehicle(MPV) intelligent safety and defence system
CN107539276A (en) * 2017-07-04 2018-01-05 芜湖市振华戎科智能科技有限公司 Car man carrier interconnected system based on CNN
CN108357462A (en) * 2018-02-09 2018-08-03 北京车和家信息技术有限公司 The method and vehicle of vehicle unlock
CN109059947A (en) * 2018-07-04 2018-12-21 深圳明创自控技术有限公司 A kind of high performance onboard navigation system
CN109693644A (en) * 2018-11-26 2019-04-30 北京建筑大学 A kind of VATS Vehicle Anti-Theft System and method with multi-biological characteristic authentication
CN110027509A (en) * 2019-04-10 2019-07-19 上海工程技术大学 A kind of vehicular theft-prevention monitoring device and method based on recognition of face

Also Published As

Publication number Publication date
CN100341732C (en) 2007-10-10

Similar Documents

Publication Publication Date Title
KR101309889B1 (en) Texture features for biometric authentication
US8744642B2 (en) Driver identification based on face data
US20060082439A1 (en) Distributed stand-off ID verification compatible with multiple face recognition systems (FRS)
US20080212849A1 (en) Method and Apparatus For Facial Image Acquisition and Recognition
Qadri et al. Automatic number plate recognition system for vehicle identification using optical character recognition
US20040052418A1 (en) Method and apparatus for probabilistic image analysis
Trivedi et al. Looking-in and looking-out of a vehicle: Computer-vision-based enhanced vehicle safety
US20180345909A1 (en) Vehicle with wearable for identifying one or more vehicle occupants
US6498970B2 (en) Automatic access to an automobile via biometrics
US10088846B2 (en) System and method for intended passenger detection
US7362210B2 (en) System and method for gate access control
US8880291B2 (en) Methods and systems for preventing unauthorized vehicle operation using face recognition
EP1454285B1 (en) Video monitoring system with object masking
US8675925B2 (en) Spoof detection for biometric authentication
JP5360491B2 (en) Multifunctional camera system
US8483450B1 (en) Quality metrics for biometric authentication
BRPI0613023A2 (en) Methods and Arrangement for Carrying out Driver Identity Verification
EP1661063A2 (en) System and method for dynamic stand-off biometric verification
WO2004055302A1 (en) Onboard device control system, onboard device controller, and onboard device control method
CN100468443C (en) Method for accurately recognizing high speed mobile vehicle mark based on video
WO2007140988A1 (en) Method for identifying a person and acquisition device
EP1705085B1 (en) Apparatus for authenticating vehicle driver
JP2002288670A (en) Personal authentication device using facial image
CN101950355A (en) Method for detecting fatigue state of driver based on digital video
CN104013414B (en) A kind of Study in Driver Fatigue State Surveillance System based on intelligent movable mobile phone

Legal Events

Date Code Title Description
C06 Publication
C10 Request of examination as to substance
C14 Granted
C41 Transfer of the right of patent application or the patent right
COR Bibliographic change or correction in the description

Free format text: CORRECT: ADDRESS; FROM: 201203 BUILDING 6, NO.115, LANE 572, BIBO ROAD, ZHANGJIANG HI-TECH PARK, SHANGHAI CITY TO: 215311 TOWER ABCD, 12/F, SOUTH BUILDING, KUNSHANPUDONG SOFTWARE PARK, BACHENG TOWN, KUNSHAN CITY

ASS Succession or assignment of patent right

Owner name: KUNSHAN JADE TECHNOLOGIES CO., LTD.

Free format text: FORMER OWNER: SHANGHAI JADE TECHNOLOGIES CO., LTD.

Effective date: 20100719

C17 Cessation of patent right