System and Method for Vehicle Security

FIELD OF THE INVENTION

[001] The present invention relates to a system and a method for vehicle security and more particularly, to a system and method for a vehicle for providing access to authorised users of the vehicle after authenticating the user(s).

BACKGROUND OF THE INVENTION

[002] Unauthorized use of a vehicle includes, for example, a vehicle theft, a child taking a parent's vehicle without permission, or a valet taking the vehicle for a “joy ride.” There are vehicle security systems and solutions in the art to prevent such unauthorized use of the vehicle. Key-based unlocking has its own cons as loss of key or loss of battery in key fob is common. Further, in fleet of vehicles, unlocking seamlessly is a need and key based unlocking is not possible.

[003] It is also known in the art to use biometric authentication for hassle free authentication of users of the vehicle. For example, the biometric recognition systems use fingerprint information of the user to identify and authenticate the user, but fingerprints may be stolen. Besides, there are known vehicle keyless entry systems which require NFC based cards, Bluetooth based OTP authentication of the vehicle from mobile app and the like. However, all these require some hardware to be carried by the user.

[004] It is also known in the art to provide a face recognition system which may be used to identify or verify a user from a stored image or video. However, ambient lighting conditions is a critical aspect for capturing an image. Such systems may not work when the ambient light conditions are not sufficient to capture a recognizable image.

[005] Further, camera-based authentication requires the user to be directly facing the camera and it may not be able to detect faces from the sides especially before and till the user sits on the vehicle. Also, it is desired to enable detection of a face even when the face of the user is only partially visible, especially in today’s scenarios with a mask, a helmet, a visor, and the like. Besides, it may not always be possible for the user/rider to sit on the vehicle seat for getting a face front captured especially when the handlebar is locked.

[006] Alternatively, it has been suggested in the art to use a camera on a personal device of the user such as a mobile phone, a PDA device and the like, but these solutions require the user to carry the personal device all the time. Besides, if the battery of the personal device is low or the network is not available, even the authorized user/owner of the vehicle may not be able to access the vehicle which becomes frustrating.

[007] Thus, there is a need in the art for a system and method for providing and limiting access to the vehicle by pre-authorised users only, which address at least the aforementioned problems.

SUMMARY OF THE INVENTION

[008] In one aspect, the present invention is directed at a vehicle security system for a vehicle comprising a camera module, a face authenticating module, a memory module, and a vehicle accessing module. The camera module is configured for capturing face data pertaining to a detected face and communicating the face data to the face authenticating module. The face authenticating module is configured for comparing the face data of the detected face with registered faces of users of the vehicle preauthorised for access to the vehicle. The memory module is configured for storing the face data pertaining to the registered faces of the preauthorised users. The vehicle accessing module is configured to allow access to the vehicle based upon the comparison by the face authenticating module.

[009] In an embodiment, the vehicle security system comprises a registering module for registering a face of a new user. The camera module captures the face data and communicates the face data of the new user to the memory module for the new user to be the preauthorised user of the vehicle.

[010] In an embodiment, the face authenticating module is configured to store and compare the face data using data using data driven models such as an MTCNN (Multi-task Cascaded Convolutional Neural Network) model and the like.

[011] In an embodiment, the vehicle security system comprises an ambient light sensor for sensing and communicating ambient light conditions to the camera module. The camera module has an infrared module to capture the face data of the detected face depending upon the ambient light conditions. The vehicle security system has an image capturing module for capturing an image in a viewing range of the camera module and a face classification module. The face classification module classifies the detected face based upon the face data captured into a face with a mask, a face with no-mask a face with the mask and a helmet, and a face with the mask, the helmet and a visor, and the like.

[012] In an embodiment, the vehicle security system has a display module (170) located in a dashboard of the vehicle to display the result of the comparison between the detected face and the registered faces, i.e., to display whether access has been granted or not.

[013] In an embodiment, the face authenticating module has a transformation matrix generation module. The transformation matrix generation module generates a transformation matrix to transform the detected face into a front face. The face authenticating module is configured for comparing the face data of the front face with the registered faces.

[014] In an embodiment, the camera module has a facial feature extraction module. The facial feature extraction module is configured to extract key facial features from the detected face captured by the camera module.

[015] In another aspect, the invention is directed towards a method for providing access to a vehicle. The method comprises: capturing and communicating, by a camera module, face data pertaining to a detected face; comparing, by a face authenticating module, the face data pertaining to the detected face and registered faces; authenticating, by a face authenticating module, the detected face when the face data of the detected face matches with at least one of the registered faces; and providing access of the vehicle to the user based upon the authentication.

[016] In an embodiment, the method comprises registering a new user by capturing, by the camera module, the face data of the new user; and communicating the face data of the new user to the memory module.

[017] In an embodiment, the method comprises sensing, and communicating the ambient light conditions to the camera module. The capturing of the face data of the detected face is performed using an infrared module and image capturing module of the camera module depending upon the ambient light conditions.

[018] In an embodiment, the method comprises classifying the detected face into a face with mask, a face with no-mask, a face with mask and helmet; and a face with mask, helmet and visor, and the like. The classification is done based upon the face data captured. [019] In an embodiment, the method comprises displaying, by a display module, the result of the comparison. The result includes information regarding providing of access to the vehicle.

[020] In an embodiment, the method comprises comparing the detected face with an average face, generating a transformation matrix based upon the comparison, and transforming the detected face into a front face using the transformation matrix; by using a transformation matrix generation module.

[021] In an embodiment, the method comprises extracting, by a facial feature extraction module, key facial features from the detected face captured by the camera module.

BRIEF DESCRIPTION OF THE DRAWINGS

[022] Reference will be made to embodiments of the invention, examples of which may be illustrated in accompanying figures. These figures are intended to be illustrative, not limiting. Although the invention is generally described in context of these embodiments, it should be understood that it is not intended to limit the scope of the invention to these particular embodiments.

Figure 1 illustrates a vehicle security system for a vehicle, in accordance with an embodiment of the invention.

Figure 2A and Figure 2B are flow charts illustrating methods for providing access to the vehicle, in accordance with various embodiments of the invention.

Figure 3 is an illustration of an exemplary embodiment of the method and the system for providing access to the vehicle, in accordance with an embodiment of the invention.

Figure 4 is a flow chart illustrating an exemplary embodiment of the method for providing access to the vehicle, in accordance with an embodiment of the invention. DETAILED DESCRIPTION OF THE INVENTION

[023] The present invention relates to vehicle security systems and methods. More particularly, the present invention relates to such systems and methods for providing access to the vehicle using face recognition.

[024] Figure 1 illustrates a vehicle security system 100 for a vehicle 190, in accordance with an embodiment of the invention. In an embodiment, the system 100 is a part of the vehicle 190 and is integral to the vehicle 190. In another embodiment, the system 100 may be provided as embedded on an external detachable modular device. In various embodiments the vehicle 190 may be an Electric Vehicle, a Hybrid Electric Vehicle, an Internal Combustion Engine vehicle and the like. The system 100 comprises a camera module 110, a face authenticating module 120, a memory module 130, and a vehicle accessing module 140. The camera module 110 is configured for capturing face data pertaining to a detected face and communicating the face data to the face authenticating module 120. The face authenticating module 120 is configured for comparing the face data of the detected face with registered faces of users of the vehicle 190 preauthorised for access to the vehicle 190. The memory module 130 is configured for storing the face data pertaining to the registered faces of the preauthorised users. The vehicle accessing module 140 is configured to allow access to the vehicle 190 based upon the comparison by the face authenticating module 120. The vehicle accessing module 140 allows access to the vehicle 190 if the detected face matches with at least one of the registered faces / faces of preauthorized users of the vehicle 190.

[025] In an embodiment, the vehicle security system 100 comprises an ambient light sensor 160 for sensing and communicating ambient light conditions to the camera module 110. The camera module 110 has an infrared module 112 to capture the face data of the detected face depending upon the ambient light conditions. For example, the infrared module 112 is provided as an I R (InfraRed) light source from an LED (Light emitting Diode) provided on an instrument cluster of the vehicle 190. The IR light is directed towards the face of a user/rider. [026] In an embodiment, the vehicle security system 100 comprises a registering module 150 for registering a face of a new user. The camera module 110 captures the face data and communicates the face data of the new user to the memory module 130 for the new user to be the preauthorised user of the vehicle 190. The face data may be in the form of a matrix identifying and defining the key features of the face. The face data includes data pertaining to identified features of the face including dimensions of polygons formed from the lines connecting the identified features, such as length of the sides of polygons, angle between the sides of the polygons. The face data includes distance between the key features of the face and the size, color, and curvature of the key features of the face.

[027] For example, if the vehicle 190 is being used by a group of people such as members of a family or a fleet, the registering module 150 stores the registered faces of all the members authorized to use the vehicle 190. These registered faces are then stored in the memory module 130 in the form of the face data pertaining to the registered faces.

[028] In an embodiment, the vehicle security system 100 has an image capturing module 118 for capturing an image in a viewing range of the camera module 110 and a face classification module 114. The image capturing module 118 includes a digital camera, a VGA camera and the like, provided on a handlebar of the vehicle 190. The face classification module 114 classifies the detected face based upon the face data captured in to: a side face, a face with a mask, a face with no-mask a face with the mask and a helmet, and a face with the mask, the helmet and a visor of the helmet. [029] In an embodiment, the face authenticating module 120 has a transformation matrix generation module 122. The transformation matrix generation module 122 generates a transformation matrix based upon a comparison between the detected face and an average face. The transformation matrix so generated is used to transform the detected face into a front face. The face authenticating module 120 is configured for comparing the face data of the front face with the registered faces.

[030] Average faces may be specific to different regions, different genders, different classes such as day-time average faces, nighttime average faces, faces with mask, without mask, with helmet, without helmet, etc. Types of average faces include Average Face (No Mask) with low light & daytime, Average Face with mask with low light & daytime, Average Face with helmet without visor with low light & daytime, Average Face with mask and helmet (Without visor) with low light & daytime and the like.

[031] The image capturing module 118 may capture an image, which may be a face image or something other than a face. For example, the image captured may be just a tree or a bird flying past the viewing range of the image capturing module 118. Further, the detected image may be a face but may not be a front face and rather a side-face or face with a mask or helmet and the like. The side face is an indirect, non-frontal view of the face from one of the sides, left side, right side or from below when only some or partial key features of the face are visible/detected. Such a detected image/face is matched with the average face, and based upon the match the transformation matrix is generated for transforming the detected face to a front face.

[032] In an embodiment, the transformation matrix is applied on a detected face to get the front face. This includes scaling of the face, shearing, translation, rotation actions and the like to transform the detected face to a front face which the face authenticating module uses for the comparison with the registered faces.

[033] In an embodiment, the camera module 1 10 has a facial feature extraction module 116. The facial feature extraction module 116 is configured to extract key facial features from the detected face captured by the camera module 110. For example, the key facial features include - eyes, mouth, eyebrows, lips, forehead, chin, ears, nose and the like. The facial feature extraction module 116 extracts the face data pertaining to such key features from the image captured by the image capturing module 118.

[034] In an embodiment, the face authenticating module 120 is configured to store and compare the face data using data driven models such as an MTCNN (Multi-task Cascaded Convolutional Neural Network) model and the like. The face classification module 114, the facial feature extraction module 116, and the transformation matrix generation module 122 have machine learning algorithms such as data driven models - MTCNN.

[035] In an embodiment, the vehicle security system 100 has a display module 170 located in a dashboard of the vehicle 190 to display the result of the comparison between the detected face and the registered faces, i.e., to display whether access has been granted or not. For example, the display module 170 is the instrument cluster of the vehicle 190 and may communicate the result of the comparison by displaying a related image or text - “access granted”, “access denied” and the like.

[036] Figure 2A is a flow chart illustrating a method 200 for providing access to the vehicle 190, in accordance with an embodiment of the invention. The method 200 comprises: capturing and communicating (220), by a camera module 110, face data pertaining to a detected face; comparing (230), by a face authenticating module 120, the face data pertaining to the detected face and registered faces; authenticating (240), by a face authenticating module 120, the detected face when the face data of the detected face matches with at least one of the registered faces; and providing access (250) of the vehicle 190 to the user based upon the authentication 240.

[037] Figure 2B is a flow chart illustrating a method 200B for providing access to the vehicle 190, in accordance with an embodiment of the invention. The method comprises registering (210) a new user by capturing (212), by the camera module 110, the face data of the new user; and communicating (214) the face data of the new user to the memory module 130.

[038] For example, if the vehicle 190 is being used by a group of people such as members of a family or fleet, the method 200B captures the face data of all the members authorized to use the vehicle 190 and communicates the data to the memory module 130 for storing the face data and for registering the registered faces.

[039] In an embodiment, the method comprises sensing, and communicating (222) the ambient light conditions to the camera module 110. The capturing (220) of the face data of the detected face is performed using an infrared module 112 and image capturing module 118 of the camera module 110 depending upon the ambient light conditions. For example, the infrared module 112 is provided as an I R light source from the LED on the instrument cluster of the vehicle 190. The IR light is directed towards the face of the user/rider.

[040] In an embodiment, the method 200B comprises classifying (232) the detected face into a face with mask, a face with no-mask, a face with mask and helmet; and a face with mask, helmet and visor. The classification (232) is done based upon the face data captured.

[041] Types of faces include Average Face No Mask with low light & daytime, Average Face with mask with low light & daytime, Average Face with helmet without visor with low light & daytime, Average Face with mask and helmet Without visor with low light & daytime and the like. [042] In an embodiment, the method comprises displaying (252), by a display module 170, the result of the comparison (230). The result includes information regarding providing of access (250) to the vehicle 190.

[043] In an embodiment, the method comprises comparing (234) the detected face with an average face, generating (236) a transformation matrix based upon the comparison, and transforming (238) the detected face into a front face using the transformation matrix; by using a transformation matrix generation module 122.

[044] In an embodiment, the method comprises extracting, by a facial feature extraction module 116, key facial features from the detected face captured by the camera module 110.

[045] Figure 3 is an exemplary embodiment illustrating the system 100 and the method 200, in accordance with an embodiment of the invention. When the image capturing module 118 captures an input image 302 and detects a face 304, the face classification module 114 classifies the detected face 304 into a classified face 306 which is one of the categories depending upon the face data of the detected face 304. Subsequently the face feature extraction module 116 extracts the face data pertaining to key features 306kf of the classified face 306. The transformation matrix generation module 122 generates a transformation matrix 310 based upon the key features 306kf of the classified face 306 and uses the transformation matrix 310 so generated to convert the classified face 306 into a front face 320.

[046] The front face 320 is then compared with the face data pertaining to the registered faces using the face authenticating module 140 and in case of a match, access is granted to the vehicle 190. In another exemplary embodiment, the data driven model such as MTCNN is used for storing and matching the face data pertaining to the classified/detected/registered faces and the face data matching may not be an exact match but may be configured for substantial matching of the detected face and the registered face. Examples of Face Recognition Algorithms are Facenet, Mobile Facenet, OpenFace, Face recognition Python based and the like.

[047] Figure 4 illustrates an exemplary embodiment of the method 200, in accordance with an embodiment of the invention. The ambient light condition is checked (402-404). Depending upon the ambient light conditions, an image of the detected face is captured using IR or without IR (405, 407). Subsequently, the detected face is classified as low light, normal light, with mask, without mask and the like (411 ,412). A set of key features are extracted from the detected face (406, 408, 410). The extracted key features / keypoints is matched with an average face (413) and subsequently, a transformation matrix is estimated. The transformation matrix is generated for the detected face based upon the extracted key features (414,415). The transformation matrix so generated is applied to the detected face to transform the detected face into a front face. The front face is compared/matched with a dataset of registered preauthorised users and depending upon a match, vehicle access is allowed (417-419). Although the access to the vehicle is granted, the vehicle may still not start because of other constraints which are configured at factory level such as side stand still ON, brake not applied and the like (420, 421 ).

[048] Advantageously, the systems and methods provided by the invention enable highest vehicle security since stealing of fingerprints, keys, access cards and the like is ruled out as it is based upon face authentication of the preauthorised users. The access to the vehicle becomes more convenient and hassle free as the user need not carry any extra device or hardware such as a smart phone or a PDA. Even installation of application on the personal device or on the vehicle is obviated since the processing of matching the detected face with registered face is performed by and within the vehicle. [049] Further, the invention works during bad ambient light conditions with the incorporation of IR. The user is not required to remove mask, helmet and the like, or face the camera directly for recognition. Further, the user may not be sitting on the rider seat to directly face the camera. The invention works even when only a side face of the user is detected. The transformation matrix and average face comparison enable face detection and authorised vehicle access for such scenarios.

[050] Besides, the invention also provides for registering of new users in real-time and no separate, special system or device is required apart from the system and the method of the invention to enable new user registration in case of addition of new drivers/users/riders or change in vehicle ownership.

[051] While the present invention has been described with respect to certain embodiments, it will be apparent to those skilled in the art that various changes and modification may be made without departing from the scope of the invention as defined in the following claims.

REFERENCE NUMERAL LIST:

VEHICLE 190

SYSTEM 100

CAMERA MODULE 110

INFRARED MODULE 112

FACE CLASSIFICATION MODULE 114

FACIAL FEATURE EXTRACTION

MODULE 116

IMAGE CAPTURING MODULE 118

FACE AUTHENTICATING MODULE 120 TRANSFORMATION MATRIX

GENERATION MODULE 122

MEMORY MODULE 130

VEHICLE ACCESSING MODULE 140 REGISTERING MODULE 150

AMBIENT LIGHT SENSOR 160

DISPLAY MODULE 170