JP2023167964A - Signal processing device and signal processing method - Google Patents

Abstract

To provide a signal processing device which can accurately detect a minute signal.SOLUTION: A signal processing device 100 comprises: a signal amplification unit 1; a plurality of threshold response units Hm; a signal generation unit 51; a threshold calculation unit 52; and a threshold setting unit 7. The signal amplification unit 1 amplifies an input signal SG1 including a noise component n(t) and a signal component s(t) and outputs an amplification signal SGA. Each threshold response unit Hm outputs a response signal B[m] indicating a comparison result between the amplification signal SGA and a threshold signal Vt[m]. The signal generation unit 51 generates a signal value sga corresponding to the signal component s(t) on the basis of the plurality of thresholds Vt[m] and the plurality of response signals B[m]. The threshold calculation unit 52 calculates a plurality of new thresholds Vt#[m#] on the basis of a threshold function F(m#). The threshold setting unit 7 updates the plurality of thresholds Vt[m] to the plurality of new thresholds Vt#[m#]. The threshold function F(m#) is determined on the basis of the signal value sga corresponding to the signal component s(t).SELECTED DRAWING: Figure 1

Description

The present invention relates to a signal processing device and a signal processing method.

The minute signal detection device (water leakage detection device) described in Patent Document 1 includes a water leakage sound collecting plate, a first pickup sensor, a second pickup sensor, a differential amplifier, a filter circuit, and an integrated amplifier. Equipped with The first pickup sensor detects water leakage sound and outputs a minute sound detection signal. The second pickup sensor is placed near the first pickup sensor. The second pickup sensor detects water leakage sound with a sensitivity different from that of the first pickup sensor and outputs a minute sound detection signal. The differential amplification amplifier differentially amplifies the micro-sound detection signal output by the first pickup sensor and the micro-sound detection signal output by the second pickup sensor. As a result, noise can be significantly reduced. The filter circuit cuts specific frequency components (noise components) in the minute sound detection signal from the differential amplifier. Then, the filter circuit outputs the minute sound detection signal with the specific frequency component cut to the integrated amplifier. The filter circuit is a bandpass filter. The integrated amplifier amplifies the minute sound detection signal from the filter circuit to a predetermined level. Then, the integrated amplifier outputs the amplified minute sound detection signal.

Japanese Patent Application Publication No. 2011-99740

However, in the small signal detection device described in Patent Document 1, for example, the bandpass filter used as the filter circuit has a wide range of frequency selectivity, and both the differential amplifier and the filter circuit are subject to random noise. It is difficult to sufficiently reduce noise due to the fact that it is a source of noise. Therefore, it is difficult to detect minute signals with high accuracy.

The present invention has been made in view of the above problems, and an object thereof is to provide a signal processing device and a signal processing method that can detect minute signals with high accuracy.

According to one aspect of the present invention, a signal processing device includes a signal amplification section, a plurality of threshold response sections, a signal generation section, a threshold calculation section, and a threshold setting section. The signal amplification section amplifies an input signal including a noise component and a signal component, and outputs an amplified signal. The amplified signal is input to the plurality of threshold response units, and a plurality of different threshold signals are respectively input to the plurality of threshold response units. Each of the threshold response units outputs a response signal indicating a comparison result between the amplified signal and the threshold signal. The signal generation unit generates a signal value corresponding to the signal component based on the plurality of threshold values respectively set to the plurality of threshold signals and the plurality of response signals. The threshold calculation unit calculates a plurality of new thresholds to be respectively set for the plurality of threshold signals based on a threshold function. The threshold value setting unit updates the plurality of threshold values used when generating the signal value corresponding to the signal component to the new plurality of threshold values. The threshold function is determined based on the signal value corresponding to the signal component, and indicates a relationship between a threshold that can be set for the threshold signal and an identification value for identifying the threshold signal. .

In one aspect of the present invention, it is preferable that the appearance of the noise component is represented by a normal distribution. Preferably, the threshold function is a function corresponding to the characteristics of the normal distribution.

In one aspect of the present invention, it is preferable that the threshold function includes a first function and a second function. Preferably, the first function is determined based on a minimum threshold that can be set to the threshold signal and the signal value corresponding to the signal component. Preferably, the second function is determined based on a maximum threshold that can be set to the threshold signal and the signal value corresponding to the signal component.

In one aspect of the present invention, the first function is a quadratic function whose maximum value is the signal value corresponding to the signal component, and is represented by a monotonically increasing interval from the minimum threshold value to the maximum value. is preferred. Preferably, the second function is a quadratic function whose minimum value is the signal value corresponding to the signal component, and is represented by a monotonically increasing section from the minimum value to the maximum threshold.

In one aspect of the present invention, each of the plurality of response signals preferably indicates the comparison result at either a first level or a second level. The signal generation unit calculates the proportion of the first level in each of the plurality of response signals for each predetermined period arranged on the time axis, and calculates the proportion of the first level in each of the plurality of response signals. Preferably, the signal value corresponding to the signal component is generated for each predetermined period based on an approximation function that approximates a relationship between a level ratio and the plurality of threshold signals. It is preferable that the threshold calculation unit updates the threshold function every predetermined period and calculates the new plurality of thresholds based on the updated threshold function. Preferably, the threshold value setting unit updates the plurality of threshold values used when generating the signal value corresponding to the signal component to the new plurality of threshold values at each of the predetermined periods.

According to another aspect of the invention, a signal processing method includes the steps of: amplifying an input signal including a noise component and a signal component; and outputting an amplified signal; and comparing the amplified signal with each of a plurality of different threshold signals. a step of outputting a plurality of response signals each indicating a plurality of comparison results; and a step of outputting a plurality of response signals respectively indicating a plurality of comparison results, and a step of outputting a plurality of response signals based on a plurality of threshold values respectively set for the plurality of threshold signals and a plurality of response signals. a step of generating a signal value corresponding to the signal component; a step of calculating a plurality of new threshold values to be respectively set for the plurality of threshold signals based on a threshold function; and generating the signal value corresponding to the signal component. updating the plurality of threshold values used in the process to the new plurality of threshold values. The threshold function is determined based on the signal value corresponding to the signal component, and indicates a relationship between a threshold that can be set for the threshold signal and an identification value for identifying the threshold signal. .

According to the present invention, minute signals can be detected with high accuracy.

FIG. 1 is a block diagram showing a signal processing device according to an embodiment of the present invention. FIG. 3 is a diagram showing an amplified signal, a threshold value, and a response signal according to the present embodiment. FIG. 6 is a diagram schematically showing the relationship between the ratio of ON signals and the identification value of the threshold signal according to the present embodiment. It is a graph showing an approximation function concerning this embodiment. 7 is a graph showing a threshold function according to the present embodiment. 7 is a graph showing another threshold function according to the present embodiment. 3 is a flowchart showing a signal processing method according to the present embodiment. It is a flowchart showing "processing to generate an output signal" according to the present embodiment. It is a flowchart which shows "processing to calculate a plurality of new threshold values" concerning this embodiment. FIG. 1 is a diagram showing an example of a signal processing device according to the present embodiment.

Embodiments of the present invention will be described below with reference to the drawings. In addition, in the drawings, the same or corresponding parts are given the same reference numerals and the description will not be repeated.

A signal processing device 100 according to an embodiment of the present invention will be described with reference to FIGS. 1 to 6. FIG. 1 is a block diagram showing a signal processing device 100 according to this embodiment. As shown in FIG. 1, the signal processing device 100 processes an input signal SG1(t) and outputs an output signal SG2(t). Hereinafter, in order to simplify the description, the input signal SG1(t) and the output signal SG2(t) will be referred to as an input signal SG1 and an output signal SG2, respectively. "t" indicates time. Input signal SG1 includes a signal component s(t) and a noise component n(t). The noise component n(t) is white normal noise (Gaussian white noise). In other words, the appearance of the noise component n(t) is represented by a normal distribution.

Specifically, the signal processing device 100 includes a signal amplifying section 1, a threshold processing section 3, a digital processing section 5, and a threshold setting section 7.

The signal amplification section 1 amplifies the input signal SG1 and outputs the amplified signal SGA to the threshold processing section 3. Specifically, the signal amplifying section 1 has an amplification factor A, and amplifies the input signal SG1 by a factor of A. Therefore, the amplified signal SGA includes a signal component A.s(t) and a noise component A.s(t). The signal amplification section 1 is, for example, an amplification circuit including an operational amplifier and the like.

The threshold setting section 7 outputs a plurality of different threshold signals Vt[0] to Vt[M] to the threshold processing section 3. In this embodiment, "M" represents an integer of 1 or more. Hereinafter, unless it is necessary to explain the threshold signals Vt[0] to Vt[M] separately, the threshold signals Vt[0] to Vt[M] will be collectively referred to as "threshold signal Vt[m]". do. "m" is an identification value for identifying the threshold signal Vt[m]. In this embodiment, "m" is an integer greater than or equal to 0 and less than or equal to M. Note that the threshold signal Vt[m] may be referred to as a threshold signal Vt. The threshold value setting unit 7 includes, for example, a D/A converter (digital-to-analog conversion circuit), or a power supply and a resistive voltage divider circuit.

The amplified signal SGA and the threshold signal Vt[m] are input to the threshold processing unit 3. A threshold value Vt[m] is set in the threshold value signal Vt[m]. In this embodiment, in order to facilitate understanding, the same reference numeral "Vt[m]" is given to the threshold signal and the threshold value set in the threshold signal. “m” is also an identification value for identifying the threshold value Vt[m]. The threshold value Vt[m] is a voltage value. Specifically, a plurality of different threshold signals Vt[0] to Vt[M] are input to the threshold processing unit 3. A plurality of threshold values Vt[0] to Vt[M] are set to the plurality of different threshold signals Vt[0] to Vt[M], respectively. The identification value m indicates the magnitude relationship between the threshold values Vt[0] to Vt[M]. In this embodiment, the threshold value Vt[m] increases as the identification value m increases (Vt[0]<Vt[1]<...<Vt[M]). Note that the threshold value Vt[m] may be referred to as a threshold value Vt.

The threshold processing unit 3 generates a plurality of response signals B[0] to B[M] by comparing the amplified signal SGA and a plurality of threshold signals Vt[0] to Vt[M]. Then, the threshold processing section 3 outputs a plurality of response signals B[0] to B[M] to the digital processing section 5. Hereinafter, if there is no need to explain the response signals B[0] to B[M] separately, the response signals B[0] to B[M] will be collectively referred to as "response signal B[m]". do.

Specifically, the threshold processing section 3 includes a plurality of threshold response sections H0 to HM. A plurality of different threshold signals Vt[0] to Vt[M] are inputted to the plurality of threshold response units H0 to HM from the threshold value setting unit 7, respectively. Further, the amplified signal SGA is inputted from the signal amplification section 1 to each of the plurality of threshold response sections H0 to HM. Hereinafter, unless it is necessary to separately explain the threshold response units H0 to HM, the threshold response units H0 to HM will be collectively referred to as "threshold response unit Hm."

The threshold response unit Hm compares the amplified signal SGA and the threshold signal Vt[m] and sends a response signal B[m] indicating the comparison result between the amplified signal SGA and the threshold signal Vt[m] to the digital processing unit 5. Output. The threshold response unit Hm is, for example, a threshold response element such as a comparator. Specifically, the response signal B[m] indicates the comparison result between the amplified signal SGA and the threshold signal Vt[m] at either the first level LH or the second level LL. The first level LH and the second level LL are different. Therefore, the response signal B[m] is a binary signal.

In this embodiment, the response signal B[m] includes an ON signal at the first level LH and an OFF signal at the second level LL. In this case, the first level LH is a "high level" and the second level LL is a "low level."

Specifically, the threshold response unit Hm turns ON as the response signal B[m] when the voltage value of the amplified signal SGA is larger than the threshold Vt[m] set to the (level) threshold signal Vt[m]. Output a signal. On the other hand, the threshold response unit Hm outputs an OFF signal as the response signal B[m] when the voltage value (level) of the amplified signal SGA is equal to or less than the threshold Vt[m]. In this way, the response signal B[m] of the threshold response unit Hm depends on the threshold Vt[m]. Furthermore, since the noise component n(t) is superimposed on the signal component s(t), the probability that the amplified signal SGA input to the threshold response unit Hm becomes larger than the threshold Vt[m] increases. That is, the threshold response unit Hm generates the response signal B[m] using stochastic resonance. Stochastic resonance is a phenomenon in which the response of a nonlinear system to a minute signal is enhanced by a noise component n(t). Here, since the noise component n(t) changes randomly during the predetermined period T, the response signal B[m] changes to an ON signal or an OFF signal according to a change in the noise component n(t). The predetermined period T is a minute period (for example, a period on the order of milliseconds) in which the signal component s(t) can be considered to be substantially constant.

The digital processing unit 5 generates an output signal SG2 based on response signals B[0] to B[M] and thresholds Vt[1] to Vt[M], and also generates an output signal SG2 based on a threshold function F (m#) to be described later. new threshold values Vt[1] to Vt[M] are calculated. The digital processing unit 5 includes, for example, a microcomputer, a field programmable gate array (FPGA), or a central processing unit (CPU).

Specifically, the digital processing section 5 includes a signal generation section 51, a threshold calculation section 52, and a storage section 53. The storage unit 53 stores target information and result information of signal processing by the digital processing unit 5. The storage unit 53 is, for example, a semiconductor memory.

The signal generation unit 51 generates a plurality of threshold values Vt[1] to Vt[M] respectively set to the plurality of threshold signals Vt[1] to Vt[M], and a plurality of response signals B[0] to B. A signal value sga corresponding to the signal component s(t) of the input signal SG1 is generated based on [M]. In this embodiment, the signal value sga is a value indicating the signal component A·s(t) included in the amplified signal SGA. Then, the signal generation unit 51 divides the signal value sga by the amplification factor A to obtain a signal value sgb that is the division result. The signal value sgb is a value indicating the signal component s(t) included in the input signal SG1. Then, the signal generation unit 51 outputs an output signal SG2 indicating the signal value sgb. Output signal SG2 is preferably a digital signal. Note that the signal values sga and sgb are voltage values.

Specifically, first, the signal generation unit 51 calculates the ratio of ON signals r[0] to r[M] in the predetermined period T in each of the plurality of response signals B[0] to B[M]. The ON signal ratios r[0] to r[M] can also be read as "first level LH ratios r[0] to r[M]." Hereinafter, unless it is necessary to explain the ratios r[0] to r[M] separately, the ratios r[0] to r[M] will be collectively referred to as the ratio r[m]. Note that the ratio r[m] may be written as a ratio r. The ratio r[m] will be explained with reference to FIG. 2, taking the case of M=9 as an example.

FIG. 2 is a diagram showing a temporal change in the amplified signal SGA during a period Δt (<<T) which is even smaller than the predetermined period T, and examples of response signals B[0] to B[9] to the amplified signal SGA. be. The horizontal axis of each graph in FIG. 2 is time t, which is common. On the other hand, the vertical axis of each graph in FIG. 2 is the voltage value of each signal. Specifically, in the first row (top row) of FIG. 2, an example of the amplified signal SGA and ten threshold values Vt[0] to Vt[9] are shown. Furthermore, in the second to eleventh stages of FIG. 2, ten signals are output by comparing the amplified signal SGA of the first stage of FIG. 2 with each threshold value Vt[0] to Vt[9]. response signals B[0] to B[9] are shown.

As shown in FIG. 2, the response signal B[m] becomes an ON signal at time t when the amplified signal SGA is greater than the threshold value Vt[m], and turns OFF at the time t when the amplified signal SGA becomes equal to or less than the threshold value Vt[m]. It becomes a signal. Therefore, the smaller the threshold value Vt[m], the larger the ON signal ratio r[m]. In the example of FIG. 2, the ratio r[0] of ON signals in the response signal B[0] compared to the smallest threshold Vt[0] is the response signal compared to the other thresholds Vt[1] to Vt[9]. The ratio of ON signals in B[1] to B[9] is larger than r[1] to r[9]. Then, in order from response signal B[1], response signal B[2], response signal B[3], response signal B[4], response signal B[5], response signal B[6], response signal B[ 7], response signal B[8], and response signal B[9], the ON signal ratio r[m] decreases in this order.

Returning to FIG. 1, next, the signal generation unit 51 calculates the ratio r[m] of ON signals in each of the plurality of response signals B[m] and the plurality of threshold signals Vt[m] during the predetermined period T. A signal value sga corresponding to the signal component s(t) of the input signal SG1 is generated based on an approximation function that approximates the relationship. In this embodiment, the signal generation unit 51 generates a signal component s based on an approximation function that approximates the relationship between the ratio r[m] of a plurality of ON signals and the identification value m of a plurality of threshold signals Vt[m]. A signal value sga corresponding to (t) is generated. The approximation function will be explained with reference to FIGS. 3 and 4. Note that the approximation function can also be described as a fitting function.

FIG. 3 is a diagram showing the relationship between the ON signal ratio r[m] and the identification value m of the threshold signal Vt[m]. The horizontal axis indicates the identification value m. In the example of FIG. 3, the identification values m are equally spaced (linear) discrete values. The vertical axis indicates the ratio r[m] of ON signals. The ON signal ratio r[m] is expressed by "Rm/T". “Rm” is the cumulative time of ON periods in the response signal B[m] based on the threshold signal Vt[m] during the predetermined period T. The ON period is a period during the predetermined period T during which the response signal B[m] indicates an ON signal (a period during which the response signal B[m] indicates the first level LH).

As shown in FIG. 3, the relationship between the ON signal ratio r[m] and the discrimination value m of the threshold signal Vt[m] is roughly represented by a curve C. Specifically, the noise component n(t) of the input signal SG1 appears in a normal distribution. Therefore, the noise component A·n(t) of the amplified signal SGA also appears in a normal distribution. As a result, the curve C is represented by a function obtained by subtracting the cumulative distribution function from 1 for a probability density function indicating a normal distribution (hereinafter referred to as "inverted cumulative distribution function"). In other words, the ratio r[m] of ON signals appears along an inverted cumulative distribution function. Therefore, when the discrimination value m is assumed to be a continuous value, in the inverted cumulative distribution function shown by the curve C, the discrimination value K (random variable ) corresponds to the signal value sga of the signal component A·s(t) included in the amplified signal SGA.

Therefore, the inverted cumulative distribution function is used as an approximation function that approximates the relationship between the ratio r[m] of ON signals in the response signal B[m] and the discrimination value m of the threshold signal Vt[m]. Specifically, it is as follows.

First, the identification value m# of the threshold signal Vt#[m#] and the approximation function D(m#) will be explained. The identification value m# is a variable when the identification value m is treated as a continuous value. Therefore, identification value m# includes identification value m. Threshold signal Vt#[m#] is threshold signal Vt[m] indicated by identification value m#.

FIG. 4 is a graph showing the approximation function D(m#). As shown in FIG. 4, the approximation function D(m#) is a function that indicates the ratio r#[m#] of ON signals in a predetermined period T, and uses the identification value m# as a variable. Specifically, the approximation function D(m#) is an inverted cumulative distribution function. The ratio r#[m#] is the ratio r[m] indicated by the identification value m#. Note that the ratio r#[m#] may be written as a ratio r#, and the threshold signal Vt#[m#] may be written as a threshold signal Vt#.

Next, a method for calculating the signal value sgb of the output signal SG2 by the signal generation section 51 will be described with reference to FIGS. 1, 4, and 5. The signal generation unit 51 shown in FIG. 1 calculates the ratio r[m] of ON signals in each response signal B[m] during a predetermined period T. Then, the signal generation unit 51 performs fitting on the relationship between the plurality of ratios r[m] and the plurality of identification values m, and derives an approximation function D(m#). Specifically, as shown in FIG. 4, the approximate function D(m#) is derived by arranging the ratio r[m] on the "m#-r#" plane and performing fitting. Then, the signal generation unit 51 calculates the identification value K# when the ON signal ratio r# is 0.5 from the approximation function D(m#).

Next, the signal generation unit 51 calculates the signal value sgb of the output signal SG2 based on the identification value K# and the threshold function E(m#). FIG. 5 is a graph showing the threshold function E(m#) according to this embodiment. The horizontal axis indicates the identification value m# (continuous value) of the threshold signal Vt#[m#]. The vertical axis indicates threshold value Vt#[m#].

As shown in FIG. 5, the threshold function E(m#) is a function that indicates the threshold value Vt#[m#] and uses the identification value m# as a variable. Specifically, the threshold function E(m#) includes a threshold value Vt#[m#] that can be set for the threshold signal Vt(m) (FIG. 1), and a discrimination value m# that is a continuous value. Indicates the relationship between In the example of FIG. 5, the threshold function E(m#) is a linear function.

The signal generation unit 51 calculates the threshold value Vt#[K#] when the identification value m# is "K#" from the threshold function E(m#). In this case, the threshold value Vt#[K#] indicates the signal component A·s(t) included in the amplified signal SGA. In other words, the threshold value Vt#[K#] indicated by the discrimination value K# (FIG. 4) when the ON signal ratio r# is 0.5 is the signal component A.s(t) included in the amplified signal SGA. shows. This is because the noise component n(t) appears in a normal distribution, so the approximation function D(m#) in FIG. 4 becomes an inverted cumulative distribution function.

Then, the signal generation unit 51 divides the threshold value Vt#[K#] by the amplification factor A, and obtains a signal value sgb indicating the signal component s(t) included in the input signal SG1. The signal generation unit 51 outputs an output signal SG2 indicating the signal value sgb.

As described above with reference to FIGS. 4 and 5, the signal generation unit 51 calculates the number of threshold response units Hm by using the approximation function D (m#) and the threshold function E (m#). Measurement resolution can be improved without increasing it. However, the accuracy of the approximation function D (m#) and the threshold function E (m#) improves as the number of threshold response units Hm increases.

Here, as described with reference to FIG. 4, the identification value K# when the ON signal ratio r# is 0.5 is calculated from the approximation function D(m#). Therefore, how to select the threshold value Vt[m] to be set in the threshold signal Vt[m] (FIG. 1) affects the calculation accuracy of the identification value K# that requires approximate calculation. Therefore, by increasing "M" which is the maximum value of the discrimination value m, providing a large number of threshold values Vt[m] (multiple threshold response units Hm), and setting the threshold value Vt[m] finely, the discrimination The accuracy of calculating the value K# can be improved. However, depending on the required specifications of the signal processing device 100, it may not be preferable to provide a large number of threshold values Vt[m] and set the threshold values Vt[m] finely from the viewpoint of circuit mounting area and cost. Therefore, by preparing a circuit with the number and range of discrimination values m according to the required specifications of the signal processing device 100, and appropriately setting the increment of the threshold value Vt[m] according to the input signal SG1, the discrimination value K High accuracy can be achieved when calculating #.

Specifically, as a more preferable example of this embodiment, the digital processing unit 5 shown in FIG. ] Dynamically control.

Specifically, the threshold calculation unit 52 calculates a plurality of new threshold values Vt#[m#] to be respectively set for the plurality of threshold signals Vt[m] based on the threshold function F(m#). . Then, the threshold setting unit 7 converts the plurality of thresholds Vt[m] used when generating the signal value sga corresponding to the signal component s(t) included in the input signal SG1 into a plurality of new thresholds Vt#[ m#]. That is, the threshold value setting unit 7 updates the plurality of threshold values Vt[m] used last time to the plurality of new threshold values Vt#[m#].

In particular, the threshold function F(m#) used when calculating the new threshold Vt#[m#] is determined based on the signal value sga corresponding to the signal component s(t) of the input signal SG1. . Therefore, according to the present embodiment, a plurality of new threshold values Vt#[m#] located near the signal component s(t) of the input signal SG1 detected last time are set to a plurality of threshold signals Vt[m]. can. That is, a plurality of threshold values Vt[m] can be placed near the signal component s(t) of the input signal SG1 detected last time. As a result, the signal component s(t), which is a minute signal, can be detected with high accuracy while suppressing an increase in the number of threshold response units Hm. “Near the signal component s(t) of the input signal SG1 detected last time” specifically means “near the signal component A·s(t) of the amplified signal SGA detected last time”. This is because the signal value sga indicates the signal component A·s(t) of the amplified signal SGA. Furthermore, in this embodiment, the signal component s(t) of the input signal SG1 is detected using stochastic resonance phenomenon without using a filter circuit. Therefore, the influence of noise caused by the filter circuit can be avoided. As a result, the signal component s(t), which is a minute signal, can be detected with higher accuracy.

The details of the threshold function F(m#), the threshold calculation unit 52, and the threshold setting unit 7 will be described below.

First, details of the threshold function F(m#) will be explained. Threshold function F(m#) indicates the relationship between threshold value Vt#[m#] and identification value m#. Specifically, the threshold function F(m#) identifies the threshold value Vt#[m#] that can be set for the threshold signal Vt[m] (FIG. 1) and the threshold signal Vt[m]. The relationship between the identification value m# and the identification value m# when the identification value m is treated as a continuous value is shown. Since the discrimination value m (discrete value) is included in the discrimination value m# (continuous value), the threshold function F(m#) is the threshold value Vt#[m] that can be set for the threshold signal Vt[m]. #] and the identification value m for identifying the threshold signal Vt[m].

Note that the end of the reference code “Vt[m]” of the threshold value Vt[m] and the threshold value signal Vt[m], and the reference code “Vt#” of the threshold value Vt#[m#] and the threshold value signal Vt#[m#] Brackets [ ] may be added to the end of "[m#]" to indicate time t.

Further, the threshold function F(m#) is determined based on the signal value sga corresponding to the signal component s(t) included in the input signal SG1. Therefore, it is necessary to obtain the signal value sga before deriving the threshold function F(m#). Therefore, the threshold setting unit 7 sets the initial value Vt[m][0] to the threshold signal Vt[m] (FIG. 1). For example, the initial values Vt[1][0] to Vt[M][0] are equally spaced values (linear values). Then, similar to the method described with reference to FIGS. 4 and 5, the signal generation unit 51 generates the threshold values Vt[0][0] to Vt[M][0], which are initial values, and the response signal B[0 ] to B[M], calculate the ratio of ON signals r[0] to r[M], and derive the approximate function D(m#). Then, the signal generation unit 51 calculates the identification value K# using the approximation function D(m#). Furthermore, the signal generation unit 51 derives a threshold function E(m#) based on the threshold values Vt[0][0] to Vt[M][0], which are initial values. Further, the signal generation unit 51 obtains a threshold Vt#[m#] indicating the signal value sga from the threshold function E(m#) using the identification value K# as an argument. The signal value sga indicates a signal component A·s(t) included in the amplified signal SGA.

Furthermore, in order to derive the threshold function F(m#), the storage unit 53 stores the minimum value Vt_mn and the maximum value Vt_mx that the threshold value Vt[m][t] can take. Hereinafter, the minimum value Vt_mn will be described as the minimum threshold value Vt_mn, and the maximum value Vt_mx will be described as the maximum threshold value Vt_mx. The minimum threshold value Vt_mn is, for example, 0V, and the maximum threshold value Vt_mx is, for example, 5V.

FIG. 6 is a graph showing the threshold function F(m#). In FIG. 6, the vertical axis on the left side indicates the threshold value Vt#[m#][t] at time t, and the vertical axis on the right side indicates the threshold value Vt#[m#][t+T] at time t+T. The horizontal axis indicates the identification value m#. Hereinafter, for example, t=0.

As shown in FIG. 6, the threshold function F(m#) includes a first function F1(m#) and a second function F2(m#).

In order to derive the threshold function F (m#), the threshold calculation unit 52 first derives the first function F1 (m#) and the second function F2 (m#) in the "m#-Vt#" plane. do.

The first function F1 (m#) has a vertex of (m#, Vt#) = (K#, A・s(t)) in the "m#-Vt#" plane, and (m#, Vt#) = (0, Vt_mn) is an upwardly convex quadratic function. The first function F1 (m#) is expressed by equation (1).

The second function F2 (m#) has a vertex of (m#, Vt#) = (K#, A・s(t)) in the "m#-Vt#" plane, and (m#, Vt#) = (M, Vt_mx) is a downwardly convex quadratic function. The second function F2(m#) is expressed by equation (2).

Then, the threshold calculation unit 52 defines a threshold function F (m#) using the first function F1 (m#) and the second function F2 (m#), as shown in equation (3).

In the threshold function F(m#), the first function F1(m#) is based on the minimum threshold value Vt_mn that can be set in the threshold signal Vt[m] and the signal component s(t) included in the input signal SG1. It is determined based on the corresponding signal value sga (signal component A·s(t)). Note that in the example of FIG. 6, the threshold value Vt#[0][t] is set to the minimum threshold value Vt_mn.

Specifically, in the threshold function F(m#), the first function F1(m#) sets the signal value sga (signal component A・s(t)) corresponding to the signal component s(t) to the maximum value mx is a quadratic function. Further, the first function F1 (m#) is represented by a monotonically increasing interval (0≦m#<K#) from the minimum threshold value Vt_mn to the maximum value mx, as shown in equation (3).

On the other hand, in the threshold function F(m#), the second function F2(m#) is a maximum threshold value Vt_mx that can be set in the threshold signal Vt[m] and a signal component s(t ) is determined based on the signal value sga (signal component A·s(t)). Note that in the example of FIG. 6, the threshold value Vt#[M][t] is set to the maximum threshold value Vt_mx.

Specifically, the second function F2 (m#) is a quadratic function whose minimum value mn is the signal value sga (signal component A·s(t)) corresponding to the signal component s(t). Further, the second function F2 (m#) is represented by a monotonically increasing interval (K#≦m#≦M) from the minimum value mn to the maximum threshold value Vt_mx, as shown in equation (3).

When the threshold value calculation unit 52 derives the threshold function F(m#), the threshold value calculation unit 52 adds a discrimination value within a monotonically increasing interval (0≦m#<K#) among the discrimination values 0 to M to the first function F1(m#). By substituting 0 to p, new threshold values Vt#[0][t+T] to Vt#[p][t+T] are calculated. "p" is an integer smaller than K# and larger than 0. Further, the threshold calculation unit 52 substitutes the identification values K# to M within the monotonically increasing interval (K#≦m#≦M) among the identification values 0 to M to the second function F2 (m#), New threshold values Vt#[K#][t+T] to Vt#[M][t+T] are calculated.

As a result of the above, the threshold value calculation unit 52 obtains a plurality of new threshold values Vt#[0][t+T] to Vt#[M][t+T]. In this case, "t+T" is a new threshold value Vt#[0][t+T ]~Vt#[M][t+T] is obtained. In other words, "t+T" is a new threshold value calculated from the threshold function F(m#) based on the signal component A.s(t) calculated based on the ratio r[m] of ON signals in the predetermined period T following time t. This indicates that Vt#[0][t+T] to Vt#[M][t+T] have been acquired.

The threshold calculation unit 52 outputs threshold information TS to the threshold setting unit 7. The threshold information TS is a digital signal. The threshold information TS is information indicating new threshold values Vt#[0][t+T] to Vt#[M][t+T]. Therefore, the threshold setting unit 7 sets new threshold values Vt#[0][t+T] to Vt#[M][t+T] for the threshold signals Vt[0] to Vt[M], respectively, based on the threshold information TS. Set. Specifically, the threshold value setting unit 7 converts the digital signal threshold values Vt#[0][t+T] to Vt#[M][t+T] into analog signals, and converts the threshold values Vt#[0][t+T] to analog signals. Vt#[M][t+T] is output as an analog signal to threshold response units H0 to HM.

Furthermore, the threshold calculation unit 52 similarly calculates new plurality of thresholds Vt#[0][t+2T] to Vt#[M][t+2T], new plurality of thresholds Vt# for t+2T, t+3T, ... [0][t+3T] to Vt#[M][t+3T], . . . are obtained.

That is, the threshold calculation unit 52 updates the threshold function F(m#) every predetermined period T, and calculates new threshold values Vt#[m#] based on the updated threshold function F(m#). Calculate. By calculating a new threshold value Vt#[m#] using such an algorithm, a plurality of threshold values Vt[m] can be placed near the signal component A·s(t), and resolution can be improved. Furthermore, it is possible to easily deal with changes in the signal component A·s(t).

Then, the threshold value setting unit 7 updates the plurality of threshold values Vt[m] used when generating the signal value sga corresponding to the signal component s(t) included in the input signal SG1 every predetermined period T. Update to a plurality of threshold values Vt#[m#]. Therefore, according to the present embodiment, the plurality of threshold values Vt[m] are updated every predetermined period T for calculating the ON signal ratio r[m]. As a result, the signal component s(t), which is a minute signal, can be detected continuously and accurately.

Note that the signal generation unit 51 calculates the ratio r[m] of ON signals in the predetermined period T in each of the plurality of response signals B[m] for each predetermined period T arranged on the time axis. Then, the signal generation unit 51 generates an approximation function D(m#) that approximates the relationship between the ratio r[m] of ON signals in each of the plurality of response signals B[m] and the plurality of threshold signals Vt[m]. Based on this, a signal value sga corresponding to the signal component s(t) is generated every predetermined period T. Then, the signal generation unit 51 divides the signal value sga indicating the signal component A·s(t) by the amplification factor A every predetermined period T, thereby generating the signal value sgb indicating the signal component s(t). It is calculated and output as an output signal SG2.

As described above with reference to FIG. 1, the appearance of the noise component n(t) included in the input signal SG1 is represented by a normal distribution. As shown in FIG. 6, the threshold function F(m#) is a function corresponding to the characteristics of normal distribution. Therefore, according to this embodiment, a new threshold signal Vt#[m#] can be appropriately calculated according to the appearance probability of the noise component n(t).

Specifically, the threshold value Vt#[K#][t] (=A・s(t)) indicated by the discrimination value K# when the ON signal ratio r#[m#] is 0.5. In the vicinity, the threshold function F(m#) is nearly constant. This indicates that near the threshold Vt#[K#][t], the probability of the appearance of the signal component A.s(t) (signal component s(t)) is the highest, and the threshold function F(m# ) corresponds to the characteristics of a normal distribution. Therefore, by using the threshold function F(m#), many new thresholds Vt#[m#][t+T] are placed near the threshold Vt#[K#] (=A・s(t)). can.

Furthermore, as the discrimination value m# becomes smaller, the threshold function F(m#) converges toward m#=0, and as the discrimination value m# increases, the threshold function F(m#) converges to m#=M. It converges on the side. This indicates that the probability of appearance of the signal component A·s(t) (signal component s(t)) is small on the m#=0 side and on the m#=M side. Therefore, the threshold function F(m#) corresponds to the characteristics of a normal distribution.

Furthermore, in this embodiment, the threshold function F(m#) is configured by a combination of a first function F1(m#) and a second function F2(m#). Therefore, the threshold function F(m#) can be made to correspond to the characteristics of the normal distribution. As a result, a plurality of new threshold values Vt#[m#][t+T] that better reflect the distribution of the noise component n(t) can be calculated.

Furthermore, in this embodiment, the first function F1 (m#) and the second function F2 (m#) are quadratic functions. Therefore, since the threshold function F(m#) becomes simple, the derivation of the threshold function F(m#) can be speeded up.

Next, the signal processing method according to this embodiment will be explained with reference to FIG. 1 and FIGS. 7 to 9. The signal processing method is executed by the signal processing device 100. FIG. 7 is a flowchart showing the signal processing method according to this embodiment. As shown in FIG. 7, the signal processing method includes steps S1 to S6.

First, as shown in FIGS. 1 and 7, in step S1, the threshold setting unit 7 performs initial setting. Specifically, the threshold setting unit 7 sets the initial value Vt[m][0] to the threshold signal Vt[m]. The initial value Vt[m][0] is from Vt[0][0] to Vt[M][0]. Furthermore, the storage unit 53 stores a minimum threshold value Vt_mn and a maximum threshold value Vt_mx that the threshold value Vt[m] can take.

Next, in step S2, the signal amplifying section 1 starts amplifying the input signal SG1 with the amplification factor A. That is, the signal amplifying section 1 amplifies the input signal SG1 and outputs the amplified signal SGA to the signal generating section 51.

Next, in step S3, the plurality of threshold response units H0 to HM generate the plurality of response signals B[ 0] to B[M]. In other words, the threshold response unit Hm compares the voltage value of the amplified signal SGA and the threshold value Vt[m][0], and outputs a response signal B[m] indicating the comparison result. In other words, the threshold processing unit 3 compares each of the plurality of different threshold signals Vt[0][0] to Vt[M][0] with the amplified signal SGA, and generates a plurality of signals each indicating the plurality of comparison results. Outputs response signals B[0] to B[M].

Next, in step S4, the signal generation unit 51 generates a , generates an output signal SG2.

Next, in step S5, the threshold calculation unit 52 calculates new thresholds Vt#[0][T] to Vt#[M][T] based on the threshold function F(m#). Then, the threshold calculation unit 52 calculates new threshold values Vt#[0][T] to Vt#[M][T] for the threshold signals Vt[0][0] to Vt[M][0]. The threshold value setting unit 7 is controlled to set the threshold value.

Next, in step S6, the threshold value setting unit 7 converts the threshold values Vt[0][0] to Vt[M][0] used in steps S3 and S4 into the new threshold value Vt#[0] calculated in step S5. ][T] to Vt#[M][T].

Next, the process proceeds from step S6 to step S3. Then, in step S3, the threshold response units H0 to HM send response signals B[0] to B[ based on the amplified signal SGA and new thresholds Vt#[0][T] to Vt#[M][T]. M] is output. Next, in step S4, the signal generation unit 51 uses the new threshold values Vt#[0][T] to Vt#[M][T] and the response signals B[0] to B[M]. , generates an output signal SG2. Next, in step S5, the threshold calculation unit 52 further calculates new thresholds Vt#[0][2T] to Vt#[M][2T]. Next, in step S6, the threshold value setting unit 7 converts the threshold values Vt#[0][T] to Vt#[M][T] used in steps S3 and S4 into the new threshold value Vt# calculated in step S5. Update to [0][2T] to Vt#[M][2T].

Next, the process proceeds from step S6 to step S3. As described above, steps S3 to S6 are repeated, and the threshold value Vt[m] is dynamically updated every predetermined period T.

Next, details of step S4 in FIG. 7 will be explained with reference to FIG. FIG. 8 is a flowchart showing the "process of generating output signal SG2" in step S4 of FIG. As shown in FIG. 8, the "process of generating output signal SG2" includes steps S41 to S47.

As shown in FIGS. 1 and 8, first, in step S41, the signal generation unit 51 generates a ratio of ON signals based on threshold values Vt[0] to Vt[M] and response signals B[0] to [M]. Calculate r[0] to r[M].

Next, in step S42, the signal generation unit 51 indicates the relationship between the ON signal ratios r[0] to r[M] and the discrimination values 0 to M of the threshold signals Vt[0] to Vt[M]. An approximate function D[m#] is derived.

Next, in step S43, the signal generation unit 51 calculates the identification value K# when the ratio r# of ON signals becomes 0.5 in the approximation function D[m#].

Next, in step S44, in the first routine, the signal generation unit 51 converts the threshold value Vt#[K#] corresponding to the identification value K# when the ON signal ratio r# becomes 0.5 into a threshold value Vt#[K#]. Calculated based on E[m#] (FIG. 5). In addition, in the second and subsequent routines, the signal generation unit 51 converts the threshold value Vt#[K#] corresponding to the identification value K# when the ON signal ratio r# becomes 0.5 into the threshold value function F[m# ] (Figure 6). Threshold value Vt#[K#] is a signal value sga indicating signal component A·s(t) included in amplified signal SGA.

Next, in step S45, the signal generation unit 51 divides the threshold value Vt#[K#], that is, the signal value sga indicating the signal component A·s(t), by the amplification factor A. As a result, a signal value sgb indicating the signal component s(t) included in the input signal SG1 is obtained as the division result.

Next, in step S46, the signal generation unit 51 sets the signal value sgb, which is the division result, to the output signal SG2.

Next, in step S46, the signal generation section 51 outputs the output signal SG2. The process then proceeds to step S5 in FIG.

Next, details of step S5 in FIG. 7 will be explained with reference to FIG. 9. FIG. 9 is a flowchart showing the "process of calculating new plurality of threshold values Vt#[m#]" in step S5 of FIG. As shown in FIG. 9, the "process of calculating new plurality of threshold values Vt#[m#]" includes steps S51 to S53.

As shown in FIGS. 1 and 9, first, in step S51, the threshold value calculation unit 52 calculates the threshold value Vt#[K#] calculated in step S44, that is, the signal component A·s(t ) is derived from the threshold function F(m#) based on the signal value sga. The threshold value Vt#[K#] is a threshold value corresponding to the identification value K# when the ON signal ratio r# is 0.5.

Next, in step S52, the threshold calculation unit 52 obtains new thresholds Vt#[0] to Vt#[M] from the threshold function F(m#) using the identification values 0 to M as arguments.

Next, in step S53, the threshold calculation unit 52 causes the threshold setting unit to set new thresholds Vt#[0] to Vt#[M] for the threshold signals Vt[0] to Vt[M]. Control 7. The process then proceeds to step S6 in FIG.

Next, an example of the signal processing device 100 will be described with reference to FIGS. 7 and 10. FIG. 10 is a diagram showing an example of the signal processing device 100. As shown in FIG. 10, in the signal processing device 100, the signal amplification section 1 includes at least one inverting amplification circuit 11. In this embodiment, the signal amplification section 1 includes a plurality of inverting amplification circuits 11. A plurality of inverting amplifier circuits 11 are connected in series. The inverting amplifier circuit 11 is a transimpedance amplifier circuit. The plurality of inverting amplifier circuits 11 amplify the input signal SG1, and the last stage inverting amplifier circuit 11 outputs the amplified signal SGA.

Each of threshold response units H0 to HM includes a hysteresis comparator HC. The plurality of hysteresis comparators HC in the plurality of threshold response units H0 to HM are connected in parallel to the output terminal of the inverting amplifier circuit 11 at the last stage. Then, the inverting amplifier circuit 11 at the last stage outputs the amplified signal SGA to the first input terminal of the hysteresis comparator HC of each of the threshold response units H0 to HM.

The threshold value setting section 7 includes a D/A converter 70. The D/A converter 70 converts the digital threshold signals Vt#[0] to Vt#[M] outputted by the threshold calculation unit 52 into analog threshold signals Vt#[0] to Vt#[M].

The signal processing device 100 includes lines L0 to LM. Hereinafter, unless it is necessary to explain the lines L0 to LM separately, the lines L0 to LM will be collectively referred to as "line Lm." One end of the lines L0-LM is connected to the D/A converter 70, and the other end of the lines L0-LM is connected to the second input terminal of the hysteresis comparator HC of the threshold response units H0-HM, respectively. Therefore, the D/A converter 70 inputs the analog threshold signals Vt[0] to Vt[M] to the second input terminals of the hysteresis comparators HC of the threshold response units H0 to HM through the lines L0 to LM. .

The hysteresis comparator HC compares the amplified signal SGA and the threshold signal Vt[m] and outputs a response signal B[m] indicating the comparison result. Specifically, the hysteresis comparator HC executes the process of step S3 in FIG.

The digital processing unit 5 includes an FPGA 50. The FPGA 50 includes a signal generation section 51, a threshold calculation section 52, and a storage section 53. Signal generation section 51 includes a plurality of shift registers G1 to GM and a signal processing section 510. Hereinafter, unless it is necessary to explain the shift registers G1 to GM separately, the shift registers G1 to GM will be collectively referred to as "shift register Gm."

The shift register Gm stores the response signal B[m] in synchronization with the internal clock of the FPGA 50. The frequency of the internal clock is, for example, 100 MHz. The shift register Gm stores a response signal B[m] corresponding to a predetermined period T. The predetermined period T is, for example, on the order of milliseconds. The shift register Gm stores a new response signal B[m] and discards the oldest response signal B[m] in synchronization with the internal clock.

The signal processing unit 510 calculates the ON signal ratio r[m] by counting the number of times the response signal B[m] indicates an ON signal during the predetermined period T (step S41 in FIG. 8). Then, the signal processing unit 510 outputs an output signal SG2 in which a signal value sgb indicating the signal component s(t) of the input signal SG1 is set by executing the processing of steps S42 to S47 in FIG.

The embodiments of the present invention have been described above with reference to the drawings. However, the present invention is not limited to the above-described embodiments, and can be implemented in various forms without departing from the spirit thereof. Further, the plurality of components disclosed in the above embodiments can be modified as appropriate. For example, some of the components shown in one embodiment may be added to the components of another embodiment, or some of the components shown in one embodiment may be configured. Elements may be deleted from the embodiment.

In addition, the drawings mainly schematically show each component in order to facilitate understanding of the invention, and the thickness, length, number, spacing, etc. of each component shown in the drawings are Actual results may differ due to circumstances. Further, the configuration of each component shown in the above embodiment is an example, and is not particularly limited, and it goes without saying that various changes can be made without substantially departing from the effects of the present invention. .

(1) As described with reference to FIG. 1 and FIG. ] was output as threshold information TS. However, for example, when the threshold setting section 7 outputs the threshold signals Vt[0] to Vt[M] to the threshold response sections H0 to HM using a power supply and a resistance voltage divider circuit, the following modification is also possible. . That is, in this case, the threshold value setting section 7 applies a set voltage from the power supply to the resistive voltage divider circuit. Therefore, by controlling the set voltage, threshold signals Vt[0] to Vt[M] indicating desired voltage values can be output from the resistive voltage divider circuit. Therefore, in this case, the threshold calculation unit 52 outputs the threshold values Vt#[0][t+T] to Vt#[M][t+T] from the resistive voltage divider circuit to the threshold response units H0 to HM, respectively. Information on the set voltage is output to the threshold value setting section 7 as threshold value information TS. Note that the threshold value calculation unit 52 calculates threshold information TS indicating the set voltage based on the threshold values Vt#[0][t+T] to Vt#[M][t+T].

(2) As described with reference to FIG. 6, the threshold function F(m#) is expressed by equation (3). However, as long as the threshold function F(m#) corresponds to the characteristics of a normal distribution, the order of the threshold function F(m#) is not particularly limited, and may be another nonlinear function.

(3) The signal processing device 100 can be employed, for example, as a microcurrent detection circuit for a DNA sequencer. Further, for example, the signal processing device 100 can be employed as a minute signal detection circuit in a cell potential measurement circuit. In addition, for example, the signal processing device 100 can be employed as a detection circuit for weak biological signals. However, the use of the signal processing device 100 is not particularly limited, and can be employed, for example, as a minute signal detection circuit in a semiconductor circuit having a fine structure.

INDUSTRIAL APPLICATION This invention is suitably used for a signal processing apparatus and a signal processing method.

1 Signal amplification section 7 Threshold setting section 51 Signal generation section 52 Threshold calculation section 100 Signal processing device Hm, H0 to HM Threshold response section

Claims (6)

a signal amplification unit that amplifies an input signal including a noise component and a signal component and outputs an amplified signal;
a plurality of threshold response units each receiving the amplified signal and a plurality of different threshold signals, each outputting a response signal indicating a comparison result between the amplification signal and the threshold signal;
a signal generation unit that generates a signal value corresponding to the signal component based on a plurality of threshold values respectively set to the plurality of threshold signals and a plurality of the response signals;
a threshold calculation unit that calculates a plurality of new thresholds to be respectively set for the plurality of threshold signals based on a threshold function;
a threshold setting unit that updates the plurality of thresholds used when generating the signal value corresponding to the signal component to the new plurality of thresholds,
The threshold function is determined based on the signal value corresponding to the signal component, and indicates a relationship between a threshold that can be set for the threshold signal and an identification value for identifying the threshold signal. , signal processing device. the appearance of the noise component is represented by a normal distribution,
The signal processing device according to claim 1, wherein the threshold function is a function corresponding to characteristics of the normal distribution. The threshold function includes a first function and a second function,
The first function is determined based on a minimum threshold that can be set to the threshold signal and the signal value corresponding to the signal component,
The signal processing device according to claim 1 or 2, wherein the second function is determined based on a maximum threshold that can be set to the threshold signal and the signal value corresponding to the signal component. The first function is a quadratic function whose maximum value is the signal value corresponding to the signal component, and is represented by a monotonically increasing interval from the minimum threshold value to the maximum value,
The signal processing according to claim 3, wherein the second function is a quadratic function whose minimum value is the signal value corresponding to the signal component, and is represented by a monotonically increasing interval from the minimum value to the maximum threshold. Device. Each of the plurality of response signals indicates the comparison result at either a first level or a second level,
The signal generation unit calculates the proportion of the first level in each of the plurality of response signals for each predetermined period arranged on the time axis, and calculates the proportion of the first level in each of the plurality of response signals. Generating the signal value corresponding to the signal component for each predetermined period based on an approximation function that approximates the relationship between the level ratio and the plurality of threshold signals;
The threshold calculation unit updates the threshold function every predetermined period, and calculates the new plurality of thresholds based on the updated threshold function,
The threshold value setting unit updates the plurality of threshold values used when generating the signal value corresponding to the signal component to the new plurality of threshold values at each of the predetermined periods. 2. The signal processing device according to 2. amplifying an input signal including a noise component and a signal component and outputting an amplified signal;
comparing each of a plurality of different threshold signals and the amplified signal, and outputting a plurality of response signals each representing a plurality of comparison results;
generating a signal value corresponding to the signal component based on a plurality of threshold values respectively set for the plurality of threshold signals and the plurality of response signals;
calculating a plurality of new thresholds to be respectively set for the plurality of threshold signals based on a threshold function;
updating the plurality of threshold values used when generating the signal value corresponding to the signal component to the new plurality of threshold values,
The threshold function is determined based on the signal value corresponding to the signal component, and indicates a relationship between a threshold that can be set for the threshold signal and an identification value for identifying the threshold signal. , signal processing methods.

Images