KR100922813B1 - Apparatus and method for detecting impact sound in multichannel manner - Google Patents

Abstract

PURPOSE: An apparatus and a method for detecting impact sound in a multichannel manner are provided to determine impulsive sound by selectively using input signals of a specific microphone among two or more microphones having different sound pressure levels according to the peripheral status. CONSTITUTION: An apparatus for detecting impact sound in a multichannel manner comprises followings. Two or more microphones(121a,121b) have different sound pressure sensitivities and sound pressure levels. A memory(140) stores algorithm for processing each microphone characteristic information or sound signals inputted from each microphone and information related to the algorithm. A controller(160) controls the operation of peripheral equipment corresponding to impulsive sound in case the impulsive sound rhythm patterns of each sound signal are same or the impulsive sound rhythm patterns is same with the impulsive sound rhythm patterns stored in the memory in one specific sound signal.

Description

Apparatus and method for detecting multi-channel impact sound {APPARATUS AND METHOD FOR DETECTING IMPACT SOUND IN MULTICHANNEL MANNER}

The present invention relates to a multi-channel impact sound detection device and method for detecting impact sound generated in a wide range of sound pressure level bands using at least two microphones having different characteristics.

In general, a switch is used to turn on / off various electric devices.

The switch is used to control home appliances such as lighting, TV, and audio in a home. However, when the switch is fixed at a specific place or location, such as a light switch, the user must move to a specific place or location and then operate the switch. There was a hassle.

Accordingly, in recent years, a switch controlled by clapping (especially the number of claps) has been released, but the switch control method by clapping has a problem in that the recognition rate of clapping sounds falls in a place where there is a lot of noise. In other words, it is difficult to control the peripheral device because it is not easy to distinguish the applause from the noise.

In addition, since the applause sounds frequently occur in various situations (eg, when smiling and focusing attention) in daily life, there is a problem that the peripheral device can be controlled even by the unintentional applause.

In addition, as in the conventional technique of controlling a device by the number of claps, when there are many peripheral devices to be controlled or an operation to be controlled in one peripheral device increases, the number of claps corresponding to the number of the peripheral devices or the number of operations Since it should also be increased, it was difficult for the user to remember the number of claps corresponding to each peripheral device.

In addition, a device (e.g., a second peripheral) that operates by a large number of claps (e.g., five claps) and a device that operates by a small number of claps (e.g., three claps) When present together, when there was a clap sound for operating the first peripheral, there was a problem that the unintentional second peripheral can also be operated.

In addition, a conventional device for detecting an impact sound such as clapping requires a high performance microprocessor because a noise canceling circuit or a complex noise canceling algorithm needs to be applied in order to distinguish between the existing noise and the clapping sound. There was a problem of raising the unit price.

In addition, according to the characteristics of a microphone for detecting a shock sound, a device using one conventional microphone may have a large impact sound exceeding a sound pressure level that the microphone can detect or a minute that is less than the sound pressure level that the microphone can detect. When the impact sound was input, there was a difficulty in detecting the impact sound.

The present invention provides an apparatus and method for more accurately detecting an impact sound generated in a wide range of sound pressure level bands from ambient noise using at least two microphones.

In addition, the present invention is to provide a device and a method for more accurately detecting a large impact sound or a fine impact sound by widening the bandwidth of the impact sound that can be detected using two or more microphones having different characteristics.

In addition, the present invention is to provide an apparatus and method for determining the impact sound by selectively using the input signal of a specific microphone of two or more microphones having different sound pressure level characteristics according to the surrounding situation.

The present invention also provides an apparatus and method for more accurately detecting impact sounds generated in a wide range of sound pressure level bands by using at least two microphones having different sound pressure level characteristics.

In addition, the present invention is to provide an apparatus and method for detecting the impact sound more accurately by selectively using a signal input from a microphone having a lower sensitivity when the impact sound is relatively larger than the ambient noise.

In addition, the present invention is to provide an apparatus and method for detecting the impact sound more accurately by selectively using a signal input from a microphone having a high sensitivity when a relatively small impact sound is input than usual.

An apparatus according to an embodiment of the present invention for realizing the above object is at least two microphones having different characteristics related to sound pressure sensitivity or height of the sound pressure level, the characteristic information of each microphone, or simultaneously received from each microphone After detecting the algorithm for processing the sound signal and the memory for storing the related information and the impact sound rhythm pattern included in each sound signal received from the microphone, the impact sound rhythm pattern of each sound signal is the same, or And a control unit for controlling an operation of a peripheral device corresponding to the impact sound rhythm pattern when a shock sound rhythm pattern identical to the impact sound rhythm pattern previously stored in the memory is detected in only one sound signal.

In addition, the method related to an embodiment of the present invention for realizing the above object is to receive a sound signal from at least two or more microphones having sensitive characteristics at different sound pressure levels, respectively, and processing the impact sound among the processed sound signals When the impact sound is detected only in any one of the sound signals by determining the detected sound signal, the peripheral device corresponding to the rhythm pattern of the impact sound is controlled. When the impact sound is detected in all of the sound signals, the impact sound rhythm pattern is detected from each sound signal and compared with each other. If the impact sound rhythm patterns of the sound signals are identical to each other, the peripheral device corresponding to the impact sound rhythm pattern is detected. Control is made. When the impact sound rhythm patterns of the respective sound signals are different from each other, it is detected whether a rhythm pattern corresponding to each impact sound rhythm pattern is stored in the memory, and the impact sound rhythm patterns of the respective sound signals are different from each other, and any impact sound rhythm is different. When only the pattern matches the impact sound rhythm pattern previously stored in the memory, the peripheral device corresponding to the impact sound rhythm pattern is controlled, and the impact sound rhythm patterns of the respective sound signals are different from each other, and the shock sound rhythm patterns of all the sound signals match. When the rhythm patterns are stored in the memory, the error processing operation is performed.

The present invention configured as described above has the effect of widening the bandwidth of the shock sound that can be detected by using two or more microphones having different characteristics to detect the impact sound, so that a large impact sound or a minute impact sound can be detected more accurately.

In addition, the present invention by using at least two or more microphones having different characteristics in order to detect the impact sound, it is effective to improve the impact sound recognition rate by selectively receiving and processing the input signal of a specific microphone according to the ambient noise environment have.

In addition, the present invention has the effect of improving the misunderstanding of the impact sound by using the input signal of the microphone that is easy to detect the impact sound according to the ambient noise environment of at least two microphones having different sound pressure level characteristics in detecting the impact sound. .

Hereinafter, the apparatus and method for detecting the impact sound according to the present invention will be described. More specifically, the present invention receives an acoustic signal output through at least two microphones having different sound pressure level characteristics, and detects an impact sound by processing at least one input signal among the acoustic signals output from each microphone. The present invention relates to an apparatus and method for detecting a multi-channel impact sound.

In addition, in the present embodiment, when the impact sound is detected by using the signal output from at least one specific microphone as described above, as an application example, the rhythm pattern composed of the impact sound is detected, and according to the rhythm pattern of the impact sound, A method of controlling the operation of the peripheral device will be described together.

However, the detailed description of the well-known technology and its configuration which are determined to obscure the gist of the present invention will be omitted. In addition, in describing the present invention with reference to the drawings, components that perform the same function will be described with the same reference numerals.

The impact sound detected in the present invention means a sound generated by the impact, and means a sound generated when a sudden sudden force is applied to the object. The impact sound may be generated using a human body such as a hand, mouth, or foot, or may be generated by hitting an arbitrary object such as a ballpoint pen or a key against another object such as a floor, a wall, or a desk.

In the present invention, the microphone may use at least two microphones having different characteristics. The characteristic of the microphone includes a characteristic related to the sound pressure sensitivity or the height of the sound pressure level. In the present embodiment, when the sound pressure sensitivity is high or low, the sensitivity may be large or small. In addition, the microphone may have a high sensitivity at a specific sound pressure level, or a specific sound pressure bandwidth. Therefore, by using two or more highly sensitive microphones at different bandwidths, the sound pressure bandwidth detectable by the device of the present embodiment can be widened.

Hereinafter, in the present embodiment, when using two or more microphones having different characteristics as described above, the impact sound detection device may detect the impact sound by processing a signal output from at least one microphone. In addition, each impact sound may be detected by simultaneously processing signals input from the two or more microphones. In addition, it is possible to detect the rhythm of the impact sound detected as described above and to control the corresponding peripheral device according to the detected rhythm.

Figure 1 is a block diagram of an embodiment showing the configuration of a device for detecting the impact sound and the rhythm of the impact sound according to the present invention, the device outputs a signal (or control signal) corresponding to the detected impact sound rhythm corresponding to You can control peripherals or their specific actions.

The device 100 for detecting the impact sound and the rhythm of the impact sound according to the present invention includes a communication unit 110, an input unit 120, an output unit 130, a memory 140, an interface unit 150, and a controller 160. ) And a power supply unit 170 and the like. However, not all illustrated components are essential components, and may be implemented by more components than the illustrated components, or by fewer components.

Hereinafter, the components will be described in detail.

The communication unit 110 may include one or more components for wired or wireless communication between the device 100 and the plurality of peripheral devices 200A to 200N of the present invention. For example, the wireless internet module 111 or the short range communication module 112 may be included. The wireless internet module 111 refers to a module for wireless internet access and may be embedded or external. The short range communication module 112 refers to a module for short range communication, and includes Bluetooth, ultrasonic waves, radio frequency identification (RFID), ZigBee, infrared data association (IrDA), and ultra wideband (UWB). Can be used.

The input unit 120 is a component for receiving a sound and a key manipulation of a user, and a key for receiving a key 121 of a microphone for receiving a sound and a user's key manipulation for setting a rhythm pattern or a peripheral device corresponding to the rhythm pattern. It may include an input unit 122.

The microphone 121 receives an external sound signal by a microphone and processes the external sound signal into electrical sound data. In the present embodiment, the microphone 121 may include at least two microphones 121a and 121b having different sound pressure level characteristics. For example, the first microphone 121a may have a high sound pressure level characteristic, and the second microphone 121b may have a low sound pressure level characteristic. If more microphones are provided, the sound pressure level can be further subdivided. In addition, the sound pressure level characteristics may be configured to partially overlap.

Here, the characteristics of each microphone may be implemented by the microphone itself, or may be implemented by adding a separate electric circuit to the microphone of the same characteristic. For convenience, the method of differently implementing the sound pressure level characteristics of the microphones is not highly related to the gist of the present invention, and a description thereof will be omitted.

The key input unit 122 generates a command for the user to control the operation of the device, and includes a key pad, a dome switch, a touch pad (static pressure / capacitance), a jog wheel, a jog switch, and the like. Can be configured.

The output unit 130 is for outputting audio information for outputting an alarm or motion information or video information such as text or an image, and may include a display unit 131 and an alarm unit 132.

The display unit 131 may display information or an operation state processed by the apparatus 100 for detecting the impact sound and the rhythm of the impact sound of the present invention. For example, the operation state of the apparatus 100 may be displayed by a GUI (Graphic User Interface) method or may be displayed using a light emitting device (eg, an LED). There may be two or more display units 131 according to the configuration of the device. For example, a plurality of light emitting devices that output different colors according to an operating state (eg, a normal operating state or an error operating state) may be used. Data and control information for each operation state may be stored in the memory 140.

The alarm unit 132 outputs a sound signal by using an audio output device such as a speaker or a buzzer for event information or an operation state processed by the apparatus 100 of the present invention. The sound signal may be a simple sound effect or may be a guide message using a human voice. Examples of the event include a key input, an impact sound input, an error occurrence, and normal operation. Data and control information according to each event or operation state may be stored in the memory 140.

The memory 140 may store a program for processing and controlling the control unit 160, and input / output data (for example, information for detecting a shock sound, information about a shock sound rhythm, and corresponding to each impact sound rhythm). It can also perform a function for the temporary storage of the signal information to be output, information to be output according to the occurrence of the event or operation state. In addition, the memory 140 includes an algorithm for blocking input of a signal containing a lot of noise, or stopping processing of the signal, each of which is difficult to detect an impact sound among signals output from microphones having two or more characteristics. The microphone's characteristic information can be additionally stored.

The memory 140 may include at least one of a flash memory type, a hard disk type, various types of random access memory (RAM), and read-only memory (ROM). It may include a storage medium.

The interface unit 150 serves as an interface with all peripheral devices connected to the device 100 of the present invention. For example, a port for power connection, a port for wired / wireless data communication, a port for connecting an external display means or an input means or an actuator means, an audio input / output (I / O) port, and the like may be included.

The controller 160 typically controls the overall operation of the apparatus 100 for detecting the impact sound and the rhythm of the impact sound. For example, the peripheral apparatus or the device 100 itself controls the sound processing for detecting the impact sound from the input sound, the determination of the impact sound, the rhythm determination by the impact sound, and the determined rhythm of the impact sound. Here, the control of the device itself means switching the device to a standby mode or an active mode. The peripheral device may be controlled in the operation mode, but the peripheral device may not be controlled in the standby mode, and only a shock sound for switching to the operation mode may be recognized.

In addition, when at least two microphones are provided as in the present embodiment, the controller 160 may detect impact sounds by processing signals output from the microphones, respectively. However, when a lot of noise is contained in the signals, it is difficult to detect the impact sound, so that the input of the signal is interrupted or the processing of the signal is stopped, and the impact sound can be detected from a signal input from another microphone. That is, the impact sound is detected by selecting a signal output from a microphone having an appropriate characteristic (a characteristic that is easy to detect an impact sound) according to the ambient noise condition. As described above, by receiving and processing a sound signal including a shock sound which can be more clearly distinguished from the ambient noise, it is possible to increase the accuracy by reducing the misunderstanding of the shock sound.

The power supply unit 170 supplies power required for the operation of each component.

The various embodiments described herein may, for example, comprise separate software modules or hardware such as digital signal processors (DSPs) or programmable logic devices (PLDs) or microprocessors that perform one function or operation. Can be implemented. In this case, the software code may be stored in the memory 140 and executed by the controller 160.

Hereinafter, the operation of detecting the impact sound and the rhythm of the impact sound according to the present invention will be described in more detail with reference to the accompanying drawings.

2 is a block diagram of an embodiment showing a configuration for processing a signal output from at least two microphones in accordance with the present invention, each component shown in this may be composed of separate hardware, the controller 160 It may be configured as one component within, or may be configured as a software module to be performed by the control unit 160.

As shown in the drawing, the controller 160 may receive signals output from at least two microphones 121a and 121b through separate signal input ports. As such, when a signal output from each microphone is input through a separate signal input port, the controller 160 internally receives both signals output from both microphones through port switching, or is output from one microphone. The signal may be selectively input.

Alternatively, when the external switching unit 180 is provided, each microphone is connected to one side (input side) of the switching unit 180 and the other side (output side) is connected to a specific port of the controller 160. . In addition, the control unit 180 may output a control signal (or a switching signal) to the switching unit 180 to receive both signals output from both microphones, or to selectively receive only signals output from any one microphone.

In the present embodiment, the control unit 160 processes the signals received from the two microphones 121a and 121b in parallel, but the processing method of each signal may be the same or slightly different. Only the processing method will be described.

Figure 3 is a block diagram of an embodiment showing the configuration of a device for detecting the impact sound and the rhythm of the impact sound in the signal received from the microphone according to the present invention.

As shown in the drawing, when the sound signal is input through the microphone 121, the controller 160 samples the sound signal at a specific frequency (for example, 8 KHz, 16 KHz) through the A / D converter 161. Then, as shown in FIG. 4A, a digital value (for example, 0 to 64,000, or -32,000 to +32,000) according to the level (amplitude) of the input sound signal is converted and output. As shown, the higher the level of the signal is converted to a digital value having a larger value, and the lower the level of the signal is converted to a digital value having a smaller value.

In this case, when the sound pressure level characteristics of the microphones 121a and 121b are different, the acoustic signals of the microphones output through the A / D converter 161 are also different. For example, FIGS. 4B to 4D compare sound signals detected by a highly sensitive (= sensitive) microphone 121a and a less sensitive (= sensitive) microphone 121b at a specific sound pressure level.

As shown in FIG. 4B, when an impact sound (eg, clapping) is input in a quiet state, the sensitive microphone 121a detects the impact sound well at an appropriate level (301). However, under the same conditions, the insensitive microphone 121b is detected 302 at a relatively small level of impact sound. Therefore, in this case, the controller 160 may detect the impact sound by processing the signal output from the sensitive microphone 121a.

In addition, as shown in FIG. 4C, when a shock sound is input while a specific noise (for example, radio sound) 303 is present in the surroundings, the sensitive microphone 121a mixes the impact sound 304 with the ambient noise to distinguish applause well. (305). However, under the same conditions, the insensitive microphone 121b has a good applause (306) compared to the ambient noise. Therefore, in this case, the controller 160 may detect the impact sound by processing the signal output from the insensitive microphone 121b.

In other words, when the TV or radio is turned on or when several people are talking, when a shock sound relatively louder than the ambient noise is input, the controller 160 detects a dull micro shock sound at which the ambient noise is hardly input. Can be. At this time, the sensitive microphone, which is well inputted with the ambient noise, has a large dispersion value of the impact sound detection algorithm. Therefore, even if the maximum input value is input, the control unit 160 switches to a state in which it is not recognized as a shock sound, and thus misunderstanding due to noise is a source. Is blocked.

In addition, as shown in FIG. 4D, when a shock sound is input in a place where a sound sounds, such as a bathroom or a corridor, the sensitive microphone 121a may not be distinguished how many impact sounds are due to the surrounding sound (307). However, in the same condition, the insensitive microphone 121b is sounded, but a signal capable of distinguishing a relatively shock sound is output (308). Therefore, in this case, the controller 160 may detect the impact sound by processing the signal output from the insensitive microphone 121b.

In other words, the microphone insensitive to a place where the ringing is severe, such as a bathroom or a corridor, does not input a shock sound thickly, so that the controller 160 may improve the recognition rate of the shock sound. That is, in the case of a single channel impact sound detection method using a conventional microphone to detect the impact sound, the non-impact sound such as the voice and the impact sound such as the clap have a strong tendency to be input in a similar waveform. There was a difficulty in distinguishing between the impact sound and the sound. However, the present invention allows a single channel type impact sound detection method to lower the probability of false recognition that may occur in a noise environment. That is, the non-impact sound and the impact sound can be distinguished more accurately.

As described above, the present invention can be referred to as a multi-channel impact sound detection method in order to be compared with the conventional single-channel shock sound detection method, compared to the conventional single-channel shock sound detection method, voice (voice) in the surroundings Even when there is a non-impact noise, such as a non-impact microphone, it is possible to more accurately detect the impact sound input to a louder sound than the non-impact noise from the detected acoustic signal.

The sound frame processor 162 divides the sound signal converted into a digital value according to the level into frame units having a predetermined time interval (for example, 50 ms), and divides the sound signal having the maximum value in each frame into the sound of the corresponding frame. The peak value Pc is set. If the sound signal is unstable during the actual sound signal processing (for example, if there is ambient noise or if the power supply is unstable even if there is no ambient noise), the signal may be unstable. The difference value between may be used to detect an acoustic signal having the maximum value.

In the present exemplary embodiment, the controller 160 may control the sound frame processor 162 to sequentially process signals received from the microphones 121a and 121b. For example, the signals input from each microphone are buffered in a specific area of the memory 140, and the frames of the sound signals received from the first microphone 121a and the frames of the sound signals received from the second microphone 121b are sequentially processed. Can be treated as

In this case, since the processing time required for processing one frame by the controller 160 is about 20 ms and smaller than the size of one frame (eg 50 ms), that is, the frame processing time is a time interval of one frame (eg 50 ms). Since it is more than twice as fast, the signals received from the first and second microphones 121a and 121b can be processed during the time interval of one frame (for example, 50 ms). That is, the sound peak value Pc can be detected in the corresponding frame of the sound signal received from each microphone for about the same time (for example, within 50 ms).

However, if the performance of the control unit 160 is poor, for example, if the processing time required to process one frame in the control unit 160 is about 30ms, for a time interval of one frame (for example, 50ms Cannot process two frames, i.e., the frame processing time is not more than twice as fast as the time interval of one frame (e.g., 50 ms), so that the second microphone 121b is used while buffering the signal of the first microphone 121a. ) And alternately process the signal of the first microphone 121a while buffering the signal of the second microphone 121.

Also, as illustrated in FIG. 5A, the sound frame processing unit 162 calculates a magnitude average value Pa of the sound peak values Pc of each frame for a predetermined number of frames (for example, 100). The predetermined frames for calculating the size average value Pa are each time a new frame is input, by excluding the frame inputted in time (oldest frame) and adding the new frame (current frame), Always maintain a constant number of frames (e.g. 100). For example, the magnitude average value Pa becomes larger or smaller in proportion to the magnitude of the ambient noise.

The frame processing as described above is performed on the signals output from the microphones 121a and 121b, respectively. That is, the magnitude average value Pa of each signal output from each microphone is calculated.

Therefore, in the present invention, when the magnitude average value Pa is small, the basic setting magnification T of the reference value (detection threshold value) for detecting the impact sound is increased (e.g., 5-10 times the magnitude average value), and the magnitude average value Pa is obtained. In this case, the default setting magnification T of the reference value (detection threshold value) for the impact sound detection is made small (e.g., 2-3 times the magnitude average value). For example, as shown in FIG. 5B, when the size average value Pa is as small as 100, the default setting magnification T of the detection threshold is set to about 10, and when the size average value Pa is as large as about 500, the basic value of the detection threshold is as follows. The setting magnification T is set to about 3.

The default preset magnification T can be set in advance by statistical values by experiment. However, the basic preset magnification T is applicable to the case where the ambient noise is always the same. Actually, the noise that may be present in the surroundings is a noise N1 having a large variation and a small variation, as shown in FIG. 5C. Noise N2 may be present in various ways.

Therefore, the magnitude average values Pa of the two noises N1 and N2 may be the same, but the degree of dispersion may be different.

Therefore, the sound frame processing unit 162 calculates a dispersion average value Da in order to determine how the sound peak value Pc of the current frame is dispersed from the calculated size average value Pa. That is, the variance average value Da determines whether the type of noise distributed in the surrounding is an impulsive noise having a large variation such as noise generated in a TV or a continuous noise having a small variation such as noise generated in a refrigerator. As described above, similarly to the size average value Pa, an update is performed for each frame by averaging the variance value of each frame with respect to a predetermined number of frames (for example, 100).

In this case, to calculate the variance value of each frame, one of any known variance equations using squares and roots may be used. However, in the present invention, the magnitude average value Pa and the acoustic peak value of the current frame may be used to reduce the calculation amount. It can be replaced by the difference from (Pc).

The variance average value Da increases according to the degree of change in the amount of ambient noise, that is, when the change in the amount of ambient noise is large (for example, the degree of watching TV), and the magnitude of the ambient noise does not change. If it is nearly constant (eg, talking to the Soggongon), the value is small.

Therefore, in the present invention, the variance average value Da is reflected when the detection threshold is set. That is, a weight W (1 = 1 (Da / Pa)) reflecting a ratio of the relative ratio of the variance average value Da to the size average value Pa is calculated, and the weight W is set as the default setting magnification of the detection threshold value. The value T * W multiplied by (T) is determined as the final set magnification (Tf = T * W) of the detection threshold. That is, the weight W also becomes large when the variance average value Da is large, and the weight W becomes small when the variance average value Da is small.

For example, assuming that the magnitude average value Pa is 100 and the variance average value Da is 100, the weight W = 1+ (Da / Pa) becomes 2, but the variance average value Da is 200, which is relatively large. If it is assumed, the weight (W = 1 + (Da / Pa)) is also increased to 3. Therefore, assuming that the basic set magnification T when the weight W is not set to 10, the final set magnification Tf of the threshold for the impact sound detection is weighted when the variance average value Da is large. When applying (W = 3), it becomes 30, and when applying the weight (W = 2) when the variance average value Da is small, it becomes 20. That is, even when the magnitude average value (Pa) is the same, when the impact sound of 30 times or more of the magnitude average value (Pa) is input in the noise environment having a large change width, the impact sound is recognized. If a shock sound is input, it is recognized as a shock sound.

As described above, the present invention considers the magnitude average value Pa of the input sound and the dispersion average value Da of the sound at the same time, and the type of noise (degree of dispersion of noise) and the magnitude of the threshold for detecting the impact sound in the vicinity. By adaptively changing according to the change width, it is possible to accurately distinguish noise and impact sound according to the surrounding situation.

That is, when there is a noise (a noise having a large change in size) that is largely different from the size average value Pa in the vicinity, the final set magnification Tf of the detection threshold is set to be relatively larger, whereby the same size average value Pa is obtained. It should be recognized as a shock sound only when a relatively loud impact sound is input than when there is a noise having a small difference from the magnitude average value Pa (a constant noise with little change in magnitude). In the case where there is a noise (a constant noise with little change in size) that has little difference with the size average value Pa in the periphery, the final set magnification Tf of the detection threshold is set to be relatively smaller. It is possible to recognize the impact sound even if a relatively small impact sound is input than when there is a noise having a Pa) but a difference between the magnitude average value Pa and the noise having a large change in size.

The impact sound determination unit 163 sets a preset magnification ratio (Tf = T * W) of the final detection threshold value (Pc / Pa) obtained by dividing the acoustic peak value Pc of each frame by the magnitude average value Pa. If larger, the impact sound is judged to have started (exactly, it is determined that the sound which is likely to be the impact sound is input), and the impact sound input time is stored in a specific area of the memory 140.

For reference, assuming that there is no noise in the surroundings, the acoustic signal should not include ripple, but if the power supply is poor, the signal flow may wave like a wave, and in such a state, the acoustic peak value (Pc) Degree changes occur. Therefore, in the present invention, in order to avoid the influence of the wave phenomenon, as described above, the value Pc / Pa obtained by dividing the acoustic peak value Pc by the magnitude average value Pa is used as the acoustic peak value for determining the impact sound, and the detection threshold value is used. The impact sound is judged by comparison with the final setting magnification Tf for setting.

For example, as shown in FIG. 5D, it is assumed that the acoustic peak value Pc of the current frame is 200, the magnitude average value Pa is 10, the weight W is 1.5, and the default set magnification T without weight is 10. In this case, the final set magnification Tf is 15. Therefore, since the sound peak value Pc of the current frame divided by the magnitude average value Pa (Pc / Pa = 20) is greater than the final set magnification (Tf = 15) to which the weight (W) is applied, the impact sound input is started. To judge. When the impact sound larger than the final set magnification Tf becomes smaller near the magnitude average value Pa again within a predetermined number of predetermined frames, the sound is finally determined as the impact sound.

The time information of the frame determined as the impact sound (for example, the time of the frame determined as the impact sound, or the time interval of each frame determined as the impact sound) is sequentially stored in the specific region of the memory 140. In addition, the impact sound determination unit 163 determines that the rhythm input of the impact sound is terminated when the impact sound is no longer input for a predetermined time (for example, 500 ms) after the detection of the impact sound.

On the other hand, when the impact sound that is greater than or equal to the final set magnification Tf exceeds a predetermined number of frames, it may not be determined as a shock sound, and the number of frames (shock sound holding time) in which the shock sound may be maintained for determining the impact sound may be around. It can be set in consideration of the situation (e.g., a place with a ringing, a place without a ringing). For example, when the level of the impact sound is too large, such as the signal 307 output from the sensitive microphone 121a of FIG. 4D, the shock sound may not be determined. In this case, however, the impact sound can be determined using the signal 308 output from the insensitive microphone 121b.

The rhythm determination unit 164 generates a rhythm pattern based on the currently input impact sound based on the number of impact sounds stored in the memory 140 and time information on the impact sound, and compares the rhythm pattern with previously stored rhythm patterns. The rhythm pattern is detected by comparing the rhythm patterns within an error range (for example, 5%) among a plurality of stored rhythm patterns. For example, referring to FIG. 4A, when the number of impact sounds is five (C1 to C5) and the time intervals t1 to t4 between each impact sound are 200, 400, 200, and 400 (ms), respectively, A rhythm pattern having a number of impact sounds is detected among the rhythm patterns, and a rhythm pattern having a time interval of the impact sound among the rhythm patterns is detected within a specific error range and a time interval of the input impact sound rhythm pattern.

On the other hand, the rhythm pattern may vary in the time (speed) according to the user. For example, the time intervals t1 to t4 between the impact sounds C1 to C5 may be increased to 160, 320, 160, and 320 (ms), or may be slowed to 230, 460, 230, and 460 (ms).

Accordingly, the rhythm determination unit 164 may perform an operation of adjusting the time signature to detect the rhythm pattern. That is, by increasing or decreasing the time interval of each of the input shock sounds at a predetermined rate, after matching the time interval of the pre-stored rhythm pattern (that is, the beat of the pre-stored rhythm pattern rhythm pattern) After matching, the same rhythm pattern can be detected within a certain error range. For example, when the total time interval of the stored rhythm pattern is 1200ms (200 + 400 + 200 + 400), the total time interval (960ms = 160 + 320 + 160 + 320) of the input shock sound is 1200ms. After expansion by a predetermined ratio, the time interval between each impact sound is compared with each time interval of the pre-stored rhythm pattern to determine whether it is the same rhythm pattern within an error range.

The signal output unit 165 outputs a predetermined signal (or control signal) corresponding to the rhythm pattern detected by the rhythm determination unit 164 to control the corresponding peripheral device. Alternatively, it is possible to control the internal operation of the device 100 for detecting the impact sound and the rhythm of the impact sound according to the present invention.

The above-described components may operate under the control of the controller 160.

6A to 6E are flowcharts for explaining the impact sound and the method for detecting the rhythm of the impact sound according to the present invention.

As shown in FIGS. 6A to 6E, when the apparatus 100 according to the present invention is turned on, a sound signal is input through each of the microphones 121a and 121b, and the sound signals are converted into digital values and output. . The controller 160 processes each digital value to determine whether an impact sound is input from each sound signal. The impact sound may be input in both signals or may be input in only one signal. In this case, the controller 160 may stop the processing of the acoustic signal in which the impact sound is not input, and the time for stopping the processing may be limited until the processing of the other sound signal is completed. That is, when the processing of the other sound signal is completed, it is possible to process the sound signal received from the two microphones again.

When the impact sound is detected in at least one of the two acoustic signals as described above, the controller 160 determines the rhythm of the impact sound and outputs a control signal corresponding to the impact sound rhythm to control the peripheral device.

Hereinafter, the method will be described in more detail.

6A is a flowchart for explaining a method of calculating a basic detection threshold value according to a default magnification by a size average value as an example related to the present invention.

As shown in FIG. 6A, when sound signals are input from the two microphones 121a and 121b, respectively, the controller 160 converts each sound signal into a digital value (S102), and converts each sound signal into a frame for a predetermined time. The sound peak value Pc having the maximum digital value in each frame is detected by dividing the data into units (S103). Then, the magnitude average value Pa of the acoustic peak value Pc of each frame is calculated for the predetermined predetermined frame (S104). In addition, a basic detection threshold value according to a preset default magnification T may be calculated using the magnitude average value Pa (S105).

The default setting magnification T may be stored in advance in the memory 140 as an experience value. Each of the values (eg, Pa, Pc, and T) may be separately detected and stored for each sound signal.

However, in the present invention, in order to increase the accuracy of the impact sound detection, the detection threshold value reflecting the variance average value Da of the noise is set.

6B is a flowchart for describing a method of calculating a final setting magnification reflecting a weight according to a dispersion degree of noise in a basic setting magnification T as an example related to the present invention, and calculating a final detection threshold accordingly.

As shown in FIG. 6B, the controller 160 calculates a variance average value Da of the noise for each acoustic signal (S201), and uses the ratio of the variance average value Da to the magnitude average value Pa to determine the noise. The weight (W = 1 + (Da / Pa)) according to the change width is calculated (S202). Then, the calculated weight W is multiplied by the basic set magnification T to calculate a set magnification Tf for setting the final detection threshold (S203). That is, the final threshold for detecting the impact sound is determined by the final set magnification Tf.

The weight W = 1 + (Da / Pa) is a value in which the relative ratio Da / Pa of the variance average value Da with respect to the size average value Pa is reflected.

6C is a flowchart for describing a method of determining an impact sound using a detection threshold set by a final setting magnification as an example related to the present invention.

As shown in FIG. 6C, when the final setting magnification Tf for determining the detection threshold value is calculated for each sound signal, the controller 160 divides the sound peak value Pc of the current frame by the magnitude average value Pa. (Pc / Pa) is calculated (S301), and when the calculated value is larger than the final set magnification (Tf = T * W) ((Pc / Pa)> Tf), it is determined as an impact sound (S302, S303). In this case, the sound signal determined as the impact sound should not exceed a predetermined number of predetermined frames.

6D is a flowchart illustrating a method of outputting and controlling a predetermined control signal to a peripheral device corresponding to an input shock sound rhythm as an example related to the present invention.

Referring to FIG. 6D, the controller 160 stores the number of impact sounds determined in each sound signal and time interval information between the impact sounds in the memory 140, and the impact sound rhythm pattern based on the number of impact sounds and the time interval information. Determine (S401).

In addition, the input shock sound rhythm pattern and the impact sound rhythm pattern previously stored in the memory 140 are compared (S402), and a rhythm pattern matching within an error range is detected (S403), and corresponds to the detected rhythm pattern. The peripheral device is controlled by outputting a signal (or a control signal) (S404).

6E is a flowchart illustrating a method of detecting an impact sound rhythm when a plurality of sound signals are input as an example related to the present invention.

As shown therein, the controller 160 receives a plurality of sound signals (S501) and processes the sound signals respectively (S502). In operation S503, an acoustic signal in which an impact sound is detected is determined from each acoustic signal. If an impact sound is detected only in one of the acoustic signals (NO in S503), the rhythm pattern including the impact sound detected in the one acoustic signal is detected using the method described in FIG. The corresponding peripheral device is controlled (S512).

However, when the impact sound is detected in all of the plurality of sound signals (YES in S503), the impact sound rhythm pattern is detected in each sound signal (S504). The impact sound rhythm patterns of the respective acoustic signals are compared with each other (S505) to determine whether each impact sound rhythm pattern is the same rhythm pattern (S506). When the determined rhythm pattern of each sound signal is the same (YES in S506), the peripheral device is controlled by outputting a signal (or a control signal) corresponding to the rhythm pattern (S512). For example, when the impact sound is detected using a plurality of microphones having different characteristics as in the present invention, the number of impact sounds that can be detected in each sound signal may be different. As such, when the number of impact sounds is different, the rhythm pattern is also different. This is because the impact sound rhythm pattern detected from the sound signal output from each microphone may be different even if the sound signals input to each microphone are the same.

However, when the impact sound rhythm pattern of each of the sound signals determined is different (NO in S506), each impact sound rhythm pattern is compared with the impact sound rhythm pattern stored in the memory 140 (S507), and the rhythm is matched within the error range. It is detected whether there is a pattern (S508). If all of the rhythm patterns corresponding to the impulsive sound rhythm patterns of the respective sound signals are stored in the memory 140 (NO in S508), it is determined as an error and the preset error processing operation is performed (S509). The error processing operation outputs an error message (or alarm) through the output unit 130.

On the other hand, when the impact sound rhythm patterns of the determined sound signals are different from each other, and only one impact sound rhythm pattern matches the rhythm pattern stored in the memory 140 (YES in S508), the signal corresponding to the rhythm pattern ( Or a control signal) to control the peripheral device (S512).

Here, a plurality of impact sound rhythm patterns may be stored in the memory 140, and a plurality of impact sound rhythm patterns stored in the memory 140 may be set through the input unit 120, and then the impact sound rhythm pattern may be stored in the memory 140. In response to each impact sound rhythm pattern, a peripheral device or a specific peripheral device to be controlled may be set. In this case, a plurality of peripheral devices or operations may be simultaneously set for one rhythm pattern. In addition, the information for setting the rhythm pattern may be output through the output unit 130, or may transmit the information edited by the computer through the interface with the computer (PC) and store it in the memory 140.

As described above, the present invention uses an acoustic signal output from a microphone having a sensitive characteristic to detect an impact sound that is input as a very small sound (for example, applause of a child) in a very quiet place such as midnight. In addition, by using the acoustic signal output from the insensitive characteristic microphone, the impact sound (for example, the impact sound with a large fluctuation range) that comes in close proximity to the impact sound detection device can be detected more accurately than the conventional art.

As described above, the apparatus 100 for controlling the peripheral apparatus using the impact sound and the impact sound rhythm related to the present invention may be variously connected in various ways in a wireless or wired manner through the communication unit 110 or the interface unit 150 as shown in FIG. It may interface with the peripheral devices 200A to 200N at the same time, or may form a network with another plurality of devices 100.

In addition, depending on the rhythm pattern setting, a plurality of peripherals can be simultaneously controlled (e.g., lighting of several rooms are on / off at the same time), or a specific operation can be controlled by selecting one of a plurality of peripherals (e.g., lighting , TV, air conditioner, door lock, on / off respectively).

In the above, preferred embodiments of the present invention have been described with reference to the accompanying drawings.

Here, the terms or words used in the present specification and claims should not be construed as being limited to the common or dictionary meanings, but should be interpreted as meanings and concepts corresponding to the technical spirit of the present invention.

Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention, and do not represent all of the technical idea of the present invention, which can be replaced at the time of the present application It should be understood that there may be various equivalents and variations.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram of an embodiment showing a configuration of an apparatus for detecting an impact sound and a rhythm of the impact sound according to the present invention.

Figure 2 is a block diagram of one embodiment showing a configuration for processing signals output from at least two microphones in accordance with the present invention.

Figure 3 is a block diagram of an embodiment showing the configuration of a device for detecting the impact sound and the rhythm of the impact sound in the signal received from the microphone according to the present invention.

4A to 4D are exemplary diagrams for explaining and comparing sound signals input from a plurality of microphones having different characteristics according to the present invention.

5A to 5D are exemplary views of graphs for explaining the impact sound detection method according to the present invention.

6A is a flowchart for explaining a method of calculating a basic detection threshold value according to a default magnification by a size average value as an example related to the present invention;

6B is a flowchart illustrating a method of calculating a final setting magnification reflecting a weight according to a dispersion degree of noise in a basic setting magnification T as an example related to the present invention, and calculating a final detection threshold accordingly.

6C is a flowchart for explaining a method of determining an impact sound using a detection threshold set by a final setting magnification as an example related to the present invention;

6D is a flowchart illustrating a method of outputting and controlling a predetermined control signal to a peripheral device corresponding to an input shock sound rhythm as an example related to the present invention.

6E is a flowchart illustrating a method of detecting an impact sound rhythm when a plurality of sound signals are input as an example related to the present invention.

7 is an exemplary view for explaining a method of applying and using the apparatus according to the present invention.

Claims (8)

At least two microphones having different characteristics related to sound pressure sensitivity or level of sound pressure level; A memory for storing characteristic information of each microphone or an algorithm for processing sound signals simultaneously input from each microphone and information related thereto; After detecting the impact sound rhythm pattern included in each sound signal input from each microphone, the impact sound rhythm patterns of each sound signal are the same, or the impact sound rhythm pattern previously stored in the memory only in one sound signal. And a controller for controlling an operation of a peripheral device corresponding to the impact sound rhythm pattern when the rhythm pattern is detected. delete The method of claim 1, wherein the control unit, According to the noise condition of the surroundings, the input of the acoustic signal, which is difficult to detect the impact sound, is interrupted or the processing of the acoustic signal is stopped, and the impact sound is detected from the acoustic signal of another microphone which is easy to detect the impact sound. Multi-channel impact sound detection device. The method of claim 1, wherein the control unit, After buffering the sound signals received from the at least two microphones in a specific area of the memory, and dividing the sound signals received from each microphone into frame units of a predetermined time, one of the two or more microphones ( While buffering the sound signal of the first microphone, the sound signal of another microphone (second microphone) is processed, and conversely, the sound signal of the first microphone is alternately processed while buffering the sound signal of the second microphone. Multi-channel impact sound detection device, characterized in that. delete delete Receiving and processing an acoustic signal from at least two microphones having sensitive characteristics at different sound pressure levels; Determining an acoustic signal in which an impact sound is detected among the processed acoustic signals; Controlling a peripheral device corresponding to the rhythm pattern of the impact sound when the impact sound is detected only in one of the acoustic signals; Detecting impact sound rhythm patterns in each of the sound signals and comparing them with each other if impact sounds are detected in all of the sound signals; Controlling peripheral devices corresponding to the impact sound rhythm pattern when the impact sound rhythm patterns of the sound signals are the same; Detecting whether a rhythm pattern corresponding to each impact sound rhythm pattern is stored in a memory when the impact sound rhythm patterns of the sound signals are different from each other; Controlling peripheral devices corresponding to the impact sound rhythm pattern when the impact sound rhythm patterns of the sound signals are different from each other and only one impact sound rhythm pattern matches the impact sound rhythm pattern previously stored in the memory; And executing an error processing operation when the rhythm patterns of the sound signals that are different from each other and the rhythm patterns that match the impulsive sound rhythm patterns of all the sound signals are stored in the memory. Channel type impact sound detection method. delete

Images