| 摘要 |
<p>The present invention pertains to the field of controlling vehicles using security systems, in particular for the protection of different types of object such as vehicles from unauthorised access, e.g. in anti-hijacking systems. In essence, the proposed vehicle control system with protection from unauthorised access based on automated speech analysis comprises: a unit for identifying text-independent parameters which determines the spectrum-time characteristics of the user's voice and manner of speech; and a unit for identifying text-dependent parameters for verifying the user's identity and intention, this unit selecting a set of parameters sufficient to ensure reliable recognition of the user's pronunciation of a word or phrase. These identification units are connected via their inputs to a microphone and linked to a subsystem for verifying the user's identity and to a subsystem which recognises operating words and control commands. Such subsystems produce signals for the adaptive control unit which is connected to the logic functions memory unit, to the ignition key which activates the system, and to other vehicle subsystems. Depending on the method applied in such a system, the system is capable, once the user has taught it his or her voice, of recognising that voice even if it is altered by (for example) illness, and will actuate the appropriate vehicle alarm subsystems in the event of unauthorised attempts to gain access to the vehicle, as well as carrying out various other procedures relating to 're-training', 'valeting' and 'car-jacking'.</p> |
