CN106327813A - Intelligent voice recognition and alarm method and system thereof - Google Patents

Abstract

The present invention discloses an intelligent voice recognition and alarm method and a system thereof. The intelligent voice recognition and alarm method comprises the following steps of S1, acquiring the voice alarm information and recognizing the voice alarm information; S2, matching the recognized voice alarm information with a pre-stored alarm phrase; S3, when the recognized voice alarm information is matched with the pre-stored alarm phrase, determining the alarm information and starting the alarm mode; S4, in the alarm mode, sending an alarm signal containing the alarm information to a predetermined receiving terminal. According to the intelligent voice recognition and alarm method and the system, the alarm accuracy can be improved. Meanwhile, the alarm is sent out timely, so that the safety of lives and properties is protected.

Description

Intelligent voice recognition alarm method and system

technical field

The invention relates to an alarm device, in particular to an intelligent voice recognition alarm method and system.

Background technique

With the continuous improvement of living standards, people's safety awareness is also increasing day by day. How can the living quarters ensure the safety of residents to the greatest extent, and when the safety of residents is violated, they can get alarm information in time and reach the scene, and find ways to relieve them? has become a problem that urgently needs to be solved. We often see from the Internet that some criminals enter the house to commit crimes such as theft, robbery, and murder. Although the victim speaks out and resists vigorously, he still fails to call for rescue. When he finds out, everything is too late; It is found that there is a fire, the smoke is filled, and the person concerned has lost his way and cannot call the police; or the lonely old man with limited mobility suddenly feels unwell, and when he gets to the phone, he is out of control; all of this is because he cannot call the police. Or the tragedy caused by not reporting to the police in time.

To solve this problem, the voice alarm system has become the best choice. The voice alarm does not need to be contacted, only a few simple words need to be spoken, the alarm system can analyze and judge according to the collected voice signal content, and then the alarm information can be transmitted to the family and the community control center, so that the family or the community management center can take action according to the alarm information. appropriate rescue operations.

At present, the development of speech recognition technology has been relatively mature, especially the recognition accuracy of non-specific speech recognition systems with small and medium vocabularies is greater than 98%, and the recognition accuracy of speech recognition systems for specific persons is even higher. Because the voice recognition alarm system is mainly installed in the residential area, the object of use is a specific group of people in the family, and the vocabulary of the voice content is small, so the recognition accuracy can reach more than 99% under certain conditions. In addition, due to the development of large-scale integrated circuit technology, various complex speech recognition systems have been made into chips and put on the market. The cost performance of chips has been continuously improved, laying a solid foundation for the application and promotion of speech recognition technology.

From the analysis of the technology distribution of patent applications at home and abroad, the currently applied patents in the field of speech recognition are mainly distributed in 21 four-level IPC classifications, but the main patent applications are concentrated in feature extraction, model building, and recognition algorithms for speech recognition. In technology classifications such as language processing technology and devices, there are only a few patents distributed in other IPC classifications, and some IPC classifications have no patent applications at home and abroad. The analysis shows that the current research and development of speech recognition application technology is mainly concentrated in the field of communication, while the development of application technology in other fields is relatively small, such as building intelligent control. There are not many voice recognition technology used in the real sense of building intelligent control in China. products in . Although the existing "voice security equipment" (patent number: ZL200920108313.3) uses voice recognition technology to classify alarms and transmit alarm information to the community management center, the information fed back to the community control center after voice recognition only includes alarms. The type, time and location of the alarm, the voice content of the specific alarm and the situation on the spot have no way of knowing. Because voice is easily interfered by some factors, such as environmental noise, sound intensity, unintentional coincidence, etc., it will cause false alarms, which will bring great trouble to the monitoring and management of the community.

Contents of the invention

An embodiment of the present invention provides a voice recognition alarm method and system, which can send alarm information to a preset receiving end through the identification of alarm voice content, and can start the indoor monitoring camera to monitor and understand the alarm situation. This processing is simple and safe, and can also prevent false positives and interference.

In order to solve the above-mentioned technical problems, an embodiment of the present invention provides an intelligent voice recognition alarm method, comprising: S1. collecting voice alarm information, and identifying the voice alarm information; Matching; S3. When the recognized voice alarm information matches the pre-stored alarm segment, determine the alarm information and activate the alarm mode; S4. In the alarm mode, send the alarm signal containing the alarm information to a preset receiving end.

Preferably, the alarm information includes alarm content and/or alarm type.

Preferably, the intelligent voice recognition alarm method further includes: S5. In the alarm mode, start the monitoring equipment, collect video monitoring information and send it to a preset receiving end and/or remote server.

Preferably, in step S5, in the alarm mode, the step of starting the monitoring device is automatically executed, or the receiving end issues a command to control whether to start the video monitoring.

Preferably, in step S5, the video monitoring information is sent to a remote server, and the receiving end logs into the remote server for video monitoring.

Preferably, the receiving end and/or the remote server controls the rotation of the monitoring device to change the angle of video monitoring and collection.

Another aspect of the present invention also provides an intelligent voice recognition alarm system, including a power supply and a voice input device, including a storage device, a voice recognition device, an alarm device and a receiving end; the voice input device is used to collect voice alarm information, The storage device is used to store the pre-stored alarm segments, the speech recognition device is used to recognize the voice alarm information, and matches the recognized voice alarm information with the pre-stored alarm segments, and the alarm device is used to start the alarm mode, Send an alarm signal to the receiving end.

Preferably, the voice recognition device includes a processor and a voice chip, the processor is used to initialize the voice chip, and pre-store the voice alarm segment, and the voice chip recognizes the content of the voice alarm information, and performs an operation with the pre-stored alarm segment. Compare and confirm the alarm information.

Preferably, a monitoring device is also included, and in the start-up alarm mode, the monitoring device is started, or the receiving end controls whether to start the monitoring device.

Preferably, the monitoring device is used to collect video monitoring information, and send the video monitoring information to a preset receiving end and/or a remote server.

Preferably, the video monitoring information is sent to a remote server, and the receiving end logs into the remote server for video monitoring.

Preferably, the receiving end and/or the remote server controls the rotation of the monitoring device to change the angle of video collection.

The voice recognition alarm method and system of the present invention collect voice alarm information and identify the voice alarm information; match the recognized voice alarm information with the pre-stored alarm segment; when the recognized voice alarm information matches the pre-stored alarm segment , confirm the alarm information and start the alarm mode; in the alarm mode, send the alarm signal containing the alarm information to the preset receiver. By adopting the technical solution of the present invention, the alarm signal can be sent to the preset receiving end through the identification of the alarm voice information, and the indoor monitoring camera can be started to monitor and understand the alarm situation, and the processing can be done accordingly, which is simple and safe , also prevents false positives and interference.

Description of drawings

Fig. 1 is the schematic diagram of the intelligent voice recognition alarm system of the embodiment of the present invention;

Fig. 2 is the schematic diagram of the intelligent speech recognition alarm method of the embodiment of the present invention;

In the present invention, the reference signs are as follows:

1. Voice input device; 2. Voice recognition device; 3. Storage device; 4. Alarm device 4; 5. Remote server; 6. Receiver.

detailed description

The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments, but not as a limitation of the present invention.

FIG. 1 is a schematic diagram of an intelligent voice recognition and alarm method according to an embodiment of the present invention. As shown in Figure 1, a kind of intelligent voice recognition alarm method comprises: S1. collecting voice alarm information, and identifying the voice alarm information; S2. matching the recognized voice alarm information with the pre-stored alarm segment; S3. When the recognized voice alarm information matches the pre-stored alarm segment, determine the alarm information and start the alarm mode; S4. In the alarm mode, send the alarm signal containing the alarm information to the preset receiving terminal 6. The preset receiving terminal 6 can be a receiving device of the community property, or a third-party receiving device such as a relative's mobile phone. The pre-stored alarm segments can be set by themselves. Users can set some uncommon words to represent specific alarm information. For example, the pre-stored "red star" means "burglary". When the user calls out "red star" loudly, the voice alarm information Recognition, when the "red star" is recognized as a pre-stored alarm segment, the alarm signal representing "burglary" will be sent. For example, the system sets the alarm signal of "burglary" to A, and the system will send A to the preset receiver. terminal 6, when the receiving terminal 6 receives the alarm signal, the receiving terminal 6 processes the alarm signal, displays the alarm content corresponding to the alarm signal, and processes it by a third party. Preferably, the alarm signal can be a pre-stored alarm segment matching the recognized voice alarm information. For example, when "red star" is recognized as a pre-stored alarm segment, the "burglary" represented by the pre-stored alarm segment will be ” is directly sent to the pre-set receiver 6 in the form of text or sound. Through this method, the accuracy of the alarm can be improved, and because the third party is often a relative, it will immediately take all measures to call the police, improve the effectiveness of the alarm, and protect the safety of life and property.

The alarm information can include the alarm content, such as "burglary", and after the alarm content is determined, the alarm type can be further matched. For example, the alarm type of "burglary" can be set as "bandit police". The information includes the alarm content "burglary" and the alarm type "bandit police". Of course, in order to facilitate the third party to issue an alarm as soon as possible, the alarm information may also include alarm phone numbers and the like.

Preferably, in the alarm mode, the monitoring equipment is activated, video surveillance information is collected and sent to the preset receiving end 6 and/or the remote server 5 . The step of starting the monitoring equipment can be performed automatically in the alarm mode; the receiving end 6 can also issue a command to control whether to start the video monitoring, that is, when the third-party user of the receiving end 6 thinks that the video monitoring needs to be turned on, the receiving end 6 controls the opening For video surveillance, when third-party users think that it is not necessary to turn on the video, they do not need to turn it on, which is convenient for users to choose.

The surveillance video can be sent directly to the receiving end 6, or to the remote server 5, or at the same time. In this embodiment, the monitoring video is directly sent to the receiving end 6. In fact, the video monitoring information can also be sent to the remote server 5, and the receiving end 6 logs in to the remote server 5 for video monitoring, which can improve the ability to receive video and watch At the same time, the video can be saved on the remote server 5 for future viewing and evidence collection.

The receiving end 6 and/or the remote server 5 can control the rotation of the monitoring equipment to change the angle of video monitoring and collection.

Another aspect of the present invention also provides a kind of intelligent voice recognition alarm system, comprises power supply and voice input device 1, also comprises storage device 3, voice recognition device 2, warning device 4 and receiving end 6; Voice input device 1 is used for Gather voice alarm information, storage device 3 is used for storing the pre-stored alarm segment, voice recognition device 2 is used for identifying the voice alarm information, and the voice alarm information is matched with the pre-stored alarm segment, and the alarm device is used to send a message to the receiver in the alarm mode. Terminal 6 sends an alarm signal. The preset receiving terminal 6 can be a receiving device of the community property, or a third-party receiving device such as a relative's mobile phone. The pre-stored alarm segments can be set by themselves. Users can set some uncommon words to represent specific alarm information. For example, the pre-stored "red star" means "burglary". When the "red star" is recognized as the pre-stored alarm segment, the alarm signal representing "burglary" will be recognized. For example, the system sets the alarm signal of "burglary" to A, and the system will send A to the pre-set When the receiving end 6 receives the alarm signal, the receiving end 6 processes the alarm signal, displays the alarm content corresponding to the alarm signal, and processes it by a third party. Preferably, the alarm signal can be a pre-stored alarm segment matching the recognized voice alarm information. For example, when the "red star" is recognized as a pre-stored alarm segment, the alarm content represented by the pre-stored alarm segment will be " "Burglary" is directly sent to the preset receiving terminal 6 in the form of text or sound, and the preset matching alarm type "bandit police" can be sent to the receiving terminal 6 at the same time.

Preferably, the voice recognition device 2 includes a processor and a voice chip, the processor is used to initialize the voice chip, pre-store the alarm segment, the voice chip recognizes the content of the voice alarm information, and compares it with the pre-stored alarm segment to determine the alarm message. Generally speaking, the voice alarm information enters the voice chip through the voice input device 1, and the content of the voice alarm information is judged through voice recognition, and compared with the pre-stored alarm paragraphs, the alarm content is determined. Of course, the alarm type can also be further determined. After matching Activate alarm mode.

It also includes a monitoring device. In the start-up alarm mode, the monitoring device is started, or the receiving end 6 controls whether to start the monitoring device. The monitoring equipment is used to collect video monitoring information, and send the video monitoring information to the preset receiving end 6 or the remote server 5, or to the receiving end 6 and the remote server 5 at the same time.

In the embodiment of the present invention, the video monitoring information is sent to the receiving end 6. In fact, the video monitoring information can also be sent to the remote server 5, and the receiving end 6 logs into the remote server 5 for video monitoring.

The monitoring equipment can be rotated and controlled by the receiving end 6 and/or the remote server 5 to change the angle of video collection, so that a third party can observe the situation and determine whether the received alarm information is correct or not.

The voice recognition device of the present invention can adopt ASR technology, which is based on key word list recognition technology. The processor transmits the recognized key words into the voice chip in the form of character strings, and then the key words spoken by the user can be recognized without any recording training by the user. Voice recognition device 2 includes a processor and a voice chip. In the embodiment of the present invention, the voice chip adopts a voice recognition/voice control chip based on non-specific human voice recognition technology, and high-precision A/D and D/A interfaces are integrated on the voice chip. , No need for external auxiliary Flash and RAM, that is, voice recognition/voice control/man-machine dialogue functions can be realized. Moreover, the list of recognized keywords can be dynamically edited, that is, several groups of keywords can be pre-set in the program, and the list of keywords can be dynamically set according to different situations.

In the embodiment of the present invention, the voice input device 1 is a microphone, the monitoring device is a camera, installed on the wall to avoid blocking, and the receiving end 6 is a mobile phone.

Of course, the method and device of the present invention can not only be used for some alarms related to violations of personal safety, such as bandit police, but also can set other alarm content and alarm types at the same time. When an alarm is issued, the method and device of the present invention can send alarm information such as "discomfort" to a third party, so as to facilitate timely understanding and monitoring of the conditions at home, so as to avoid delaying the illness.

Of course, the above are preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, some improvements and modifications can be made without departing from the basic principle of the present invention, and these improvements and modifications are also regarded as the protection scope of the present invention.

Claims (12)

1. An intelligent voice recognition method for reporting to the police, is characterized in that, comprises: S1. Collect voice alarm information and identify the voice alarm information; S2. Matching the recognized voice alarm information with the pre-stored alarm segment; S3. When the recognized voice alarm information matches the pre-stored alarm segment, determine the alarm information and start the alarm mode; S4. In the alarm mode, send the alarm signal including the alarm information to a preset receiving end. 2. The intelligent voice recognition alarm method according to claim 1, wherein the alarm information includes alarm content and/or alarm type. 3. intelligent voice recognition alarm method according to claim 1, is characterized in that, described intelligent voice recognition alarm method also comprises: S5. In the alarm mode, start the monitoring equipment, collect video monitoring information and send it to the preset receiving end and/or remote server. 4. The intelligent voice recognition alarm method according to claim 3, characterized in that, in the step S5, in the alarm mode, the step of starting the monitoring device is automatically executed, or the receiving end issues a command to control whether to start Video Surveillance. 5. The intelligent voice recognition alarm method according to claim 4, wherein in the step S5: the monitoring device sends the video monitoring information to a remote server, and the receiving end logs in to the remote server for Video Surveillance. 6. The intelligent voice recognition alarm method according to claim 5, characterized in that, the receiving end and/or the remote server controls the rotation of the monitoring equipment to change the angle of video monitoring and collection. 7. An intelligent voice recognition alarm system, comprising a power supply and a voice input device, is characterized in that, comprising a storage device, a voice recognition device, an alarm device and a receiving end; the voice input device is used to collect voice alarm information, and the storage The device is used to store the pre-stored alarm segment, the voice recognition device is used to recognize the voice alarm information, and matches the recognized voice alarm information with the pre-stored alarm segment, and the alarm device is used to send a message to the receiving end in the start-up alarm mode. Send an alarm signal. 8. The intelligent voice recognition alarm system according to claim 7, wherein the voice recognition device includes a processor and a voice chip, and the processor is used to initialize the voice chip, prestore the alarm segment, and the voice The chip recognizes the content of the voice alarm information and compares it with the pre-stored alarm segments to determine the alarm information. 9 . The intelligent voice recognition alarm system according to claim 8 , further comprising a monitoring device, and in the start-up alarm mode, the monitoring device is started, or the receiving end controls whether to start the monitoring device. 10. The intelligent voice recognition alarm system according to claim 9, wherein the monitoring device is used to collect video surveillance information, and send the video surveillance information to a preset receiving end and/or a remote end server. 11. The intelligent voice recognition alarm system according to claim 10, wherein the monitoring device sends video monitoring information to a remote server, and the receiving end logs into the remote server for video monitoring. 12. The intelligent voice recognition alarm system according to claim 10, characterized in that, the receiving end and/or the remote server controls the rotation of the monitoring equipment to change the angle of video collection.