WO2004102800A1 - デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム - Google Patents
デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム Download PDFInfo
- Publication number
- WO2004102800A1 WO2004102800A1 PCT/JP2004/003668 JP2004003668W WO2004102800A1 WO 2004102800 A1 WO2004102800 A1 WO 2004102800A1 JP 2004003668 W JP2004003668 W JP 2004003668W WO 2004102800 A1 WO2004102800 A1 WO 2004102800A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- filter
- filter coefficient
- coefficient
- digital
- original
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
- H03H17/00—Networks using digital techniques
- H03H17/02—Frequency selective networks
- H03H17/06—Non-recursive filters
Definitions
- the present invention relates to a digital filter, a method of designing the same, a design apparatus, and a digital filter design program, and more particularly to a design method of a FIR filter.
- FIR Finite Impulse Response
- This FIR filter is a type of filter that has a delay line with taps consisting of multiple delay devices, multiplies the output signal of each tap by a filter coefficient, adds the result of multiplication, and outputs the result. It has the following advantages. First, the pole of the transfer function of the FIR filter is only at the origin of the z-plane, so the circuit is always stable. Second, if the filter coefficients are symmetric, completely accurate linear phase characteristics can be realized.
- designing an FIR filter means determining the filter coefficients so as to obtain the desired frequency characteristics.
- a filter coefficient is calculated based on a target frequency characteristic, and windowing is performed on the filter coefficient to obtain a finite number of coefficient groups. Then, the obtained coefficient group is converted to frequency characteristics by performing FFT (Fast Fourier Transform), and this is the target characteristic. The method was designed to check whether or not the user was satisfied.
- a method of realizing a steep frequency characteristic by cascade-connecting the FIR filters generated by the above-described conventional method in a plurality of stages is known (for example, see Japanese Patent Application Laid-Open No. H5-243908).
- Also known is a method of adjusting the filter bank bandwidth by inserting a plurality of zero values between each tap (between each filter coefficient) of a tapped delay line (for example, (See Table 6 — Japanese Patent Publication No. 503450).
- the pass band of the filter worked in a direction narrower than that before the adjustment, and it was not possible to widen the bandwidth of the pass band to make the slope steeper. Disclosure of the invention
- the present invention has been made in order to solve such a problem, and a simple filter design method in which the bandwidth of the frequency characteristic is widened and the slope is steep, and an ideal frequency in which the slope is steep.
- the purpose is to provide a new FIR digital filter with characteristics.
- a digital filter design method includes a first step of generating a first filter coefficient having a symmetrical numerical sequence, and the first filter coefficient.
- a symmetrical second filter coefficient for realizing a second frequency amplitude characteristic that has a contact point at a local maximum value in the first frequency amplitude characteristic represented and realizes a second frequency amplitude characteristic at a local minimum value at the contact point is obtained.
- the digital filter of the present invention has an original filter having a first filter coefficient whose numerical sequence is symmetrical, and a contact at a position where the frequency amplitude characteristic of the original filter has a maximum value.
- An adjustment filter having a symmetrical second filter coefficient for realizing a frequency amplitude characteristic having a minimum value is provided, and the original filter and the adjustment filter are connected in cascade.
- a digital filter may be configured by a single filter set as a filter coefficient for the signal of the tap.
- an FIR digital filter having a frequency characteristic with a relatively wide passband and a steep stopband slope can be simply used without using a window function or an approximate expression. You can design.
- FIG. 1 is a flowchart showing a procedure of a method of designing a FIR digital filter according to the present embodiment.
- FIG. 2 is a frequency characteristic diagram for explaining the concept of the design method of the FIR digital filter according to the present embodiment.
- FIG. 3 is a diagram for explaining the content of the calculation for obtaining the second filter coefficient.
- FIG. 4 is a diagram showing the frequency amplitude characteristics of the original bandpass filter and the frequency amplitude characteristics obtained when one to three adjustment filters are cascaded to the original bandpass filter.
- FIG. 5 is a diagram for explaining the principle of a change in frequency characteristics obtained when the adjustment filters of the present embodiment are connected in cascade.
- FIG. 7 is a diagram showing the frequency amplitude characteristics of the original one-pass filter and the frequency amplitude characteristics obtained when one to five adjustment filters are cascaded to the original one-pass filter.
- FIG. 8 is a graph of the filter coefficient values (before rounding) actually calculated with 16-bit calculation accuracy.
- FIG. 9 is a diagram illustrating a frequency characteristic of the digital filter before the filter coefficient is rounded.
- FIG. 10 is a diagram illustrating a configuration example of the digital filter according to the present embodiment.
- FIG. 11 is a diagram showing filter coefficients of the basic port-pass filter L 4 an.
- FIG. 12 is a diagram showing a hardware configuration example of the basic port—pass filter L4a4.
- FIG. 13 is a diagram showing the frequency characteristics of the basic port-pass filter L4a4.
- FIG. 14 is a diagram showing the frequency-gain characteristics of the basic port-pass filter L 4 an.
- FIG. 15 shows the filter coefficients of the basic high-pass filter H 4 sn.
- FIG. 16 is a diagram illustrating filter coefficients of the basic bandpass filter B 4 sn. BEST MODE FOR CARRYING OUT THE INVENTION
- FIG. 1 is a flowchart showing a procedure of a method of designing a FIR digital filter according to the present embodiment.
- FIG. 2 is a frequency characteristic diagram for explaining the outline of the design method of the FIR digital filter according to the present embodiment.
- a numerical sequence generates a first filter coefficient having a symmetric type (step S 1).
- the method for generating the first filter coefficient is not particularly limited in the present invention. If the numerical sequence of filter coefficients is symmetric, a conventional design method using an approximate expression or a window function may be used.
- the design methods proposed in the past eg, Japanese Patent Application No. 2 0 0 1 — 3 6 5 1 4 6, Japanese Patent Application 2 0 0 2 — 1 2 9 16 8, Japanese Patent Application 2 0 0 2 — 2 0 5 0 1 4, No. 2
- the third design method described in Japanese Patent Application No. 2003-5662 performs an FIR operation based on a filter coefficient consisting of a predetermined basic numerical sequence, and then applies the calculated data to the calculated data.
- the filter coefficient is obtained by adding the original data before the operation and the previous data that is a predetermined amount of delay before the operation and repeating the moving average operation for adjusting the amplitude and outputting the result n times. It is preferable to use the second design method or the third design method in that good frequency characteristics of attenuation characteristics can be obtained with a small number of taps.
- a frequency characteristic indicated by a symbol A in FIG. 2 shows an example of a frequency-gain characteristic of the original filter realized by the first filter coefficient generated in step S1.
- the gain and frequency are normalized (normalized) by "1" in the frequency characteristics of the original filter.
- the frequency amplitude characteristic represented by the first filter coefficient (see FIG. 2)
- a symmetrical second filter coefficient is obtained that has a contact point at the position of the local maximum value and realizes the frequency amplitude characteristic (B in Fig. 2) that takes the local minimum value at the contact point (step S2).
- the second filter coefficient may be generated by any method, but can be obtained by, for example, the following calculation.
- the numerical sequence of the first filter coefficients that constitute the original filter is represented by ⁇ H m , H m — i ⁇ ⁇ ⁇ ⁇ H ⁇ H 0) H_ ⁇ ⁇ ⁇ , H. (m . 0 , H_ m ⁇ (H. is the median and is symmetrical about the median.
- H m H_ ra ,
- a filter having the second filter coefficient is referred to as an “adjustment filter”.
- an original filter having the first filter coefficient An adjustment having a filter coefficient of 2.
- step S3 Performing a calculation to obtain a third filter coefficient obtained when the filters are cascaded (step S3)
- the first filter coefficient and the second filter coefficient are multiplied and added to create a new filter coefficient. Be sent out.
- FIG. 3 is a diagram for explaining the contents of the calculation in step S3.
- (2m + l) (2m + l represents the number of full-value sequences forming the first filter coefficient) constituting the first filter coefficient
- the third filter is obtained. Find a numeric sequence of data coefficients.
- step S3 the third filter coefficient generated in step S3 is newly regarded as the first filter coefficient, and the process returns to step S2. Then, based on the new first filter coefficient (corresponding to a numerical sequence output from the first-stage adjustment filter when a single pulse is input to the original filter), the second filter coefficient is obtained. Find again (generate a new adjustment filter).
- a new third filter coefficient obtained when a new adjustment filter is further cascaded is obtained. Calculate. Such an operation is repeated for the number of adjustment filters to be cascaded.
- Figure 4 shows the frequency amplitude characteristics of the original filter (bandpass filter) and the frequency amplitude characteristics obtained when one to three adjustment filters are cascaded to the original filter.
- 41 is the frequency amplitude characteristic of the original filter
- 42 is the frequency amplitude characteristic obtained when one adjustment filter is cascaded
- 43 is the frequency amplitude characteristic obtained when two adjustment filters are cascaded.
- 4 shows the frequency amplitude characteristics obtained when three adjustment filters are connected in cascade.
- the adjustment filter of this embodiment is cascaded to the original filter. By connecting, the pass band width of the filter can be widened and the slope of the stop band can be made steep. Adjustment filter connected in cascade By increasing the number of filters, it is possible to obtain a filter characteristic having a wider pass bandwidth and a steeper slope.
- FIG. 5 is a diagram for explaining the principle of a change in frequency characteristics obtained when the adjustment filter of the present embodiment is connected in cascade.
- FIG. 5 (a) shows the change in the frequency amplitude characteristic when the first adjustment filter is vertically IJC-connected to the original filter.
- A is the frequency amplitude characteristic of the original filter
- B is the frequency amplitude of the first adjustment filter having the second filter coefficient generated from the first filter coefficient of the original filter.
- the characteristic C indicates the frequency amplitude characteristic obtained when the original filter / evening and the first adjustment filter are cascaded with each other.
- the new frequency amplitude characteristic C when one ⁇ filter is connected in cascade to the original filter is obtained by multiplying the frequency amplitude characteristic A of the original filter and the frequency amplitude characteristic B of the adjustment filter. Become.
- the second adjustment filter is further connected in cascade, the third filter coefficient corresponding to the frequency amplitude characteristic C generated in this manner is newly used as the first filter coefficient, and the second adjustment filter is used. New second for tuning filters Is calculated.
- FIG. 5 (b) shows a change in the frequency amplitude characteristic when the second adjustment filter is further cascaded.
- a ′ is the frequency amplitude characteristic when the first adjustment filter is cascaded, and is the same as the frequency amplitude characteristic C obtained by the procedure of FIG. 5 (a).
- B ′ is the frequency amplitude characteristic of the second adjustment filter having the new second filter coefficient generated from the new first filter coefficient corresponding to the frequency amplitude characteristic A ′.
- C ' is a new frequency amplitude characteristic obtained when the second adjustment filter is further connected in cascade, and is a form in which the two frequency amplitude characteristics A' and B 'are multiplied.
- the filter coefficient corresponding to the new frequency amplitude characteristic C ′ generated in the procedure of FIG. is obtained.
- a new second filter coefficient for the third adjustment filter is obtained.
- a new frequency amplitude characteristic is obtained according to the same procedure as described above.
- the pass band width of the filter can be increased and the slope of the stop band can be made steeper.
- k 1
- the frequency amplitude characteristic of the original filter and the frequency amplitude characteristic of the adjustment filter are line-symmetric with respect to the line whose amplitude is "1". Therefore, no matter how many adjustment filters are connected, the frequency amplitude characteristics of the new multiplied filter will not exceed the line with amplitude "1", and no overshoot or ringing will occur. For this reason, the value of k is preferably set to "1".
- the connection can be expanded by connecting one adjustment filter.
- the percentage of over-bandwidth can be increased. Therefore, when it is desired to efficiently increase the pass bandwidth with a small number of adjustment filters, it is better to increase the value of k.
- the filter coefficients are symmetric, phase linearity can be ensured.
- by adjusting the value of k as k ⁇ l it is possible to fine-tune the pass frequency bandwidth.
- FIG. 7 is a diagram showing a frequency amplitude characteristic of an original one-pass filter and a frequency amplitude characteristic obtained when 1 to 5 adjustment filters are cascaded to the original low-pass filter.
- 51 indicates the frequency amplitude characteristics of the original one-pass filter
- 52 to 56 indicate the frequency amplitude characteristics obtained when one to five adjustment filters are connected in cascade.
- the pass band width of the filter is increased and the slope of the stop band is reduced.
- the slope can be steeper. Further, by increasing the number of cascaded adjustment filters, it is possible to obtain a filter characteristic having a wider passband and a steeper slope.
- Step S4 Simplifying the Filter Coefficient by Integer Conversion (Step S4)
- Step S4 Simplifying the Filter Coefficient by Integer Conversion
- Fig. 9 is a diagram showing the frequency characteristics of the digital filter before the filter coefficients are rounded.
- A shows the gain on a linear scale and
- B shows the gain on a logarithmic scale. ing.
- the first filter coefficient is obtained by the second filter design method or the third filter design method proposed by the inventor in the past, and based on this, the second filter coefficient and the third filter coefficient are obtained.
- the value of the third filter coefficient is maximized at the center (factor H D) as shown in FIG.
- the difference between the values of the filter coefficients is much larger than that of the filter coefficients obtained by the conventional filer or metering method. Therefore, even if a filter coefficient having a value smaller than a predetermined threshold value is discarded by rounding, most of the main filter coefficients that determine the frequency characteristics remain, and the frequency characteristics are hardly affected. .
- the out-of-band attenuation of the frequency characteristic is limited by the number of bits of the filter coefficient, but as shown in FIG. 9, the frequency characteristic obtained by the second or third filter design method is very small. Deep attenuation 3 ⁇ 4: Since it can be held, even if the number of bits is slightly reduced, the desired attenuation can be secured.
- the filter can be designed without windowing, so that no truncation error occurs in the frequency characteristics. Therefore, the cutoff characteristics can be significantly improved, and excellent phase-linear filter characteristics can be obtained.
- the ripple at the flat portion is extremely small, and can be sufficiently within the range of ⁇ 0.3 dB.
- the value of the third filter coefficient obtained by the above procedure is a decimal number, and the number of digits can be reduced by rounding 10 bits, but is a set of random values.
- This numerical sequence may be used as it is as the filter coefficient, but in order to further reduce the number of multipliers used in implementing the digital filter, the numerical value of the filter coefficient is further rounded and simplified. May be.
- the numerical value sequence of the filter coefficient rounded by 10 bits is multiplied by 2 1 () , the decimal part is rounded, and the coefficient value is converted to an integer.
- integer filter coefficients can be represented by binary addition, such as 2 i + 2 i +- ⁇ • (where i and j are arbitrary integers). This makes it possible to configure a coefficient unit with a bit shift circuit instead of a multiplier, and to simplify the configuration of a digital filter to be mounted.
- the process of rounding the data of the y bits to X bits (x ⁇ y) by truncating the lower few bits from the data of the filter coefficient has been described. It is not limited to the example.
- the value of each filter coefficient may be compared with a predetermined threshold, and the filter coefficient smaller than the threshold may be discarded. In this case, the remaining filter coefficients remain at the original y bits, so when converting this to an integer, multiply by 2 y .
- An apparatus for realizing the filter design method according to the present embodiment described above can be realized by any of a hardware configuration, DSP, and software.
- the filter design apparatus of the present embodiment is actually configured by a CP / MPU, RAM, R / M, or the like of a computer, and is implemented in a RAM, a ROM, an 81-disk, or the like. This can be realized by operating the stored program.
- filter coefficients of various basic filters related to the second or third design method are stored in a storage device such as a RAM, an R ⁇ M, or a hard disk as the evening.
- the CPU uses the filter coefficient data stored in the
- the storage device corresponds to the filter coefficient storing means of the present invention, and the CPU corresponds to the calculating means of the present invention.
- the interface can be arbitrarily configured when the user indicates the combination and connection order for each basic filer, the parameters for the filter, the parameters for the filter, and the number of vertical connections for the adjustment filter. It is. For example, when the type of the basic file is displayed on the screen, it can be selected by operating the keyboard and the mouse, and the values of various parameters can be selected by operating the keyboard and the mouse.
- the CPU that can input the information obtains the information input in this way, and obtains a filer coefficient corresponding to the content specified by the input information by the above-described calculation.
- filters are iconified and displayed on the display screen (filter coefficients are stored as data corresponding to each icon in the storage device), and the user can use these icons as keyboards. Arrange in any combination on the display screen by operating the mouse or mouse. Other necessary parameters are input by keypad or mouse operation. And
- the PU calculates the fill factor coefficient corresponding to the icon array or input parameter—evening by white calculation.
- the first filter coefficient and the second filter coefficient can be obtained by using the functions of a spreadsheet software installed on a personal computer or the like.
- the calculation is actually performed by using a spreadsheet software installed. Done by ROMRAM etc.
- the obtained fill coefficient may be automatically FFT-transformed, and the result may be displayed on a display screen as a frequency-gain characteristic diagram. In this way, the frequency characteristics of the designed filter can be visually confirmed, and the filter can be more easily designed.
- Such a filter designing apparatus is realized by recording a program for causing a computer to perform the function of the present embodiment on a recording medium such as a CD-ROM, and inserting the program into a convenience store. Can appear.
- a recording medium for recording the above program a flexible disk, a hard disk, a magnetic tape, an optical disk, a magneto-optical disk, a DVD, a nonvolatile memory card, and the like can be used in addition to the CD-ROM.
- the above program can also be realized by downloading the program via a network such as the Internet at a short time.
- the functions of the above-described embodiment are realized not only by executing the supplied program, but also by the operating system (OS) or the operating system in which the program runs on the computer.
- OS operating system
- the functions of the above-described embodiment are realized in cooperation with another application software or the like, or when all or a part of the processing of the supplied program is performed by the function expansion board or function expansion board of the computer.
- Such a program is also included in the embodiment of the present invention when the functions of the above-described embodiment are implemented by a unit.
- FIG. 10 is a diagram showing a configuration example of a digital filter in that case.
- the original filter 10 and one adjustment filter 20 are connected in cascade.
- the original filter 10 is composed of i cascade-connected D-type flip-flops 11 1 to 11, (i + 1) coefficient units 12 2 —, to 1 2 _ ( i + 1 ), and i , And 13.
- the i D-type flip-flops 11 1-, to 11 delay input data sequentially by one clock CK.
- the (i + 1) coefficient multipliers 1 2-,-1 2-( m ) provide a symmetrical first signal to the signals extracted from the input and output taps of each D-type flip-flop 11-11. Multiply each filter coefficient.
- the i adders 13-, ⁇ 13 add all the multiplication results of the coefficient units 12-12 + and output the result.
- the data output from the original filter 10 is input to the adjustment filter 20 and also to the filter coefficient calculator 30.
- the filter coefficient calculator 30 calculates a second filter coefficient to be set in the adjustment filter 20 based on the output data from the original filter 10.
- the filter coefficient calculation unit 30 calculates ⁇ — k H — k H m _...,-k ⁇ ,, — k H. ten (1 + k), - k H- have ⁇ , - k H_ (B1 _ 0) - a k H.
- B ⁇ (k is arbitrarily set the positive number) second filter coefficient by comprising arithmetic determined. Then, the second fill evening coefficient determined, set respectively adjust fill evening 2 0 in the (j + 1) number of coefficient units 2 2 to 2 2 + a.
- the adjustment filter 20 is composed of j (j> i) D-type 7-lip flops 21 to 22— connected in cascade and (j + 1) coefficient units 22 ⁇ , to 22 + And j adders 23 to 23.
- j D-type flip-flops 2 1 _, ⁇ 2 1 convert the input data from the original filter 10 into 1 Delay sequentially by clock CK.
- (J + 1) coefficient units 2 2 —, ⁇ 2 2 _ (j + n are symmetrical to the signals extracted from the input and output taps of each D-type flip-flop 21-, ⁇ 21.
- the j adders 2 3 _ to 2 3 multiply by the second filter coefficients, respectively, and add and output all the multiplication results of the coefficient units 22 to 22 + .
- a digital filter may be configured so that the final filter coefficients obtained by the procedure shown in FIG. 1 are set in a plurality of coefficient units in the digital filter.
- the number of obtained filter coefficients is greatly reduced by rounding 10 bits, and is converted to a simple integer by rounding 2 1 (1 times. Therefore, the number of taps
- a multiplier is basically unnecessary and a bit shift circuit can be used, and a desired frequency characteristic can be realized with high accuracy on a small circuit scale.
- the filter design method of the present embodiment when the number of cascades of the adjustment filters is increased, the number of filter coefficients required for the subsequent adjustment filter is increased. Therefore, in order to reduce the number of taps in the entire digital filter, it is preferable to reduce the number of filter coefficients used in the original filter as a base as much as possible.
- the third design method according to Japanese Patent Application Laid-Open No. 2003-56662, filed by the inventor, an original filter having a good frequency characteristic with a very small number of sunsets is configured. It is possible to do.
- the content of Japanese Patent Application Laid-Open No. 2003-203 regarding the third design method will be extracted and described. In the following description, the parameters of m, i, j, and ⁇ ⁇ ⁇ are used, but the meaning is completely different from that described above.
- the basic filters are broadly classified into three types: basic one-pass filters, basic high-pass filters, and basic band-pass filters. Hereinafter, these basic filters will be described.
- the original data refers to the j-th data from the top of the (n-1) -th column.
- the previous data refers to the (j-1) th data from the top of the (n-1) th column.
- the first numerical value "1-1" from the top of the basic mouth one-pass filter L4a1 is obtained by adding the original data "1-1” and the previous data "0"
- the second numerical value " 3 " is obtained by adding the original data” 4 "and the previous data” -1 ".
- the third numerical value "3” is obtained by adding the former data "1 1” and the previous data "4"
- the fourth numerical value "1-1” is obtained by adding the original data "0" and the previous data "1-1”.
- the numerical sequence of "1-1, m, -1" is generated based on the original numerical sequence "-1, N".
- N 2 + ⁇ 3. That is, the coefficient of the basic unit filter is "1-1, 3.732" (here, up to three decimal places are displayed).
- the filter coefficient when the basic unit filters are connected in cascade in two stages is "1-3.732, 14.928, -3.732". This sequence of numbers has a relationship of 1: 1: 4: -1.
- the filter coefficient of the basic low-pass filter L man is equal to the sum of the numerical sequence of "1", and the total value of one skip of the numerical sequence Have the same sign and are equal to each other.
- the basic-port one-pass filter L 4 a 4 is composed of a FIR operation unit 101 having a numerical value sequence “—1/2, 2, -1/2” as a starting point as a filter coefficient.
- a moving average calculation unit 201 for performing a moving average calculation on the numerical value sequence.
- the FIR operation unit 101 includes two cascade-connected D-type flip-flops 1 to 1 _ 2 , three coefficient units 2 _ to 2 -3 , and two subtractors 3 3 _ And 2 . ⁇
- the two D-type flip-flops 1-,- 1-2 sequentially delay the input data by one clock CK.
- the three coefficient units 2— i S -s provide 1 Z 2, 2, and 1 Z 2 filter coefficients for the signals extracted from the input and output taps of each D-type flip flop 1 _, to 1 -2. Are respectively multiplied.
- First subtractor 3 the multiplication result of the second coefficient unit 2 _ 2 first coefficient unit 2 _, subtracts the multiplication result.
- the second subtracter 3 _ 2 the first subtractor 3, to the subtraction result subtracting the third coefficient unit 2 _ 3 of the multiplication result of.
- the moving average calculation unit 201 is configured by cascade-connecting four integrators 4 to 4 _ 4 each having the same configuration.
- the first stage product The divider 4 delays the input data by one clock by a D-type flip-flop 5, the original data that does not pass through the D-type flip-flop 5, and the previous data that is delayed by passing through the D-type flip-flop 5.
- an adjuster 7 for restoring the amplitude of the addition result.
- coefficient units 2 to 2 _ 3 can be constituted by a bit Toshifu bets circuit.
- the four integrators 4 - 4 - 4 comprises regulator 7 -, ⁇ 7 - 4 also can be configured in pit Toshifu bets circuit. Even if the value of n is set to a value other than 4 and the number of adjusters changes, all of the adjusters can be configured by a bit shift circuit. Therefore, in the hardware configuration of the basic low-pass filter L 4 an, no multiplier is required at all.
- FIG. 13 is a diagram showing frequency characteristics (frequency-gain characteristics and frequency-phase characteristics) obtained by performing FFT conversion on a numerical sequence of filter coefficients of the basic low-pass filter L4a4.
- the gain is shown on a linear scale, and the standardized gain is shown as 32 times.
- the frequency is standardized at "1".
- the frequency-gain characteristic has a flatness in the passband and a waiting characteristic with a gentle slope in the stopband.
- an almost linear characteristic is obtained in the frequency-phase characteristic.
- Fig. 14 shows the frequency-gain characteristics of the basic low-pass filter L4 an where n is a parameter.
- A shows the gain on a linear scale
- (b) shows the gain on a logarithmic scale.
- the slope of the stop band becomes steeper, and the bandwidth of the pass band becomes narrower.
- both ends of the top of the frequency characteristic rise.
- the original data refers to the j-th data from the top of the (n-1 1) -th column.
- ⁇ ⁇ ⁇ refers to the (j-1) th day from the top of the (n-1 1) th column.
- the largest numerical value “1” from the top of the basic eight-pass filter ⁇ 4 s 1 is obtained by subtracting the previous data “0” from the original data “1”
- the second number “3” is obtained by subtracting the previous data from the original data "4".
- the third number “One 3” is obtained by subtracting the original number "1” or bm-Eight 4 ", and the fourth number" -1
- the numerical sequence of any of the filter coefficients is symmetrical, and the total value of the numerical sequence is one with the opposite sign and equal to each other.
- the numerical sequence has a symmetric absolute value, and the first half and the second half have opposite signs.
- it has the property that the total value of one step in a numerical sequence is equal to each other with the opposite sign.
- the numerical sequence of "1, m, 1" is generated based on the original numerical sequence "1, N".
- N 2 + 3. That is, the coefficient of the basic unit filter is "1, 3.732" (three decimal places are displayed here). Also, this basic unit filter Filler coefficient for two-stage cascade connection is "3-7 3 2, 1 4.92
- the gain is normalized to 1 "by making the amplitude when FFT transform becomes" 1 ". That is, the numerical sequence of the filter coefficients actually used is" 1/2, 2, 1/2 "
- the actual numeric sequence "1 / 2,2,1 / 2" is also multiplied by X times the original numeric sequence "1,4,1".
- the filter coefficient of the basic high-pass filter Hm sn is such that the sum of the numerical sequence is "0" and the total value of the numerical sequence is one. Have the opposite sign and are equal to each other.
- the hardware configuration of the basic high-pass filter Hm sn is not shown, it is configured in the same manner as in FIG. That is, it is composed of an FIR operation unit having the output numerical sequence 1/2, 2, 2 "as a filter coefficient, and a moving average operation unit for performing a moving average operation on this numerical sequence. Even in the hardware configuration of this high-pass filter Hm sn, output taps are required only in the first-stage FIR calculation unit, and the number of evening taps is small. Also, no multiplier is required.
- the frequency characteristics of the basic high-pass filter H msn are also omitted from the figure, the frequency-gain characteristics are similar to those of the basic aperture filter L man in that the passband is almost flat and the slope of the stopband is gentle. Become.
- a linear characteristic can be obtained in the frequency-phase characteristic. In other words, a high-pass filter with good frequency characteristics without overburst You can get an evening.
- n increases, the slope of the stop band becomes steeper, and the band width of the pass band becomes narrower.
- the value of m is reduced, the slope of the stop band becomes steeper, and the bandwidth of the pass band becomes narrower.
- the moving average is calculated by sequentially subtracting the end of two BUs from the original data.
- the 7-pointer indicates the j-th data from the top of the (n-1) -th column.
- M m evening refers to the (j-2) th data from the top of the (n-1) column
- the numerical sequence of "1, 0, m, 0, 1" is generated based on the original numerical sequence "1, 0, N".
- the basic unit filter having this numerical sequence "1, 0, N” as a filter coefficient is an asymmetric type, it must be used in cascade even-numbered stages in order to make it a symmetric type.
- the numerical sequence "1, 0, N" convolution of the filter coefficient is "N, 0, N 2 + 1, 0, N".
- N (m + (m 2 ⁇ 4 ) 1/2 ) / 2.
- the gain is normalized to "1” so that the amplitude when "FF" is converted to "1". That is, the numerical sequence of the filter coefficients actually used is “1 no 2,0,2,0,1 / 2”.
- Each of the filter coefficients of the bandpass filter B msn has the property that the sum of its numerical sequence is "0" and the sum of the three values of the numerical sequence is equal to each other with the opposite sign.
- the hardware configuration of the basic bandpass filter B msn is not shown, the configuration is the same as that of FIG. In other words, it is composed of a FIR operation unit that has a numerical value sequence "1/2, 0, 2, 0, 1Z2" as a filter coefficient, and a moving average operation unit that performs a moving average operation on this numerical value sequence. . Even in the hardware configuration of the basic band-pass filter B ms n, only the first stage FIR calculation unit needs an output tap, and the number of taps is small. Also, no multiplier is required.
- the frequency characteristics of the basic band-pass filter B msn are not shown, as in the case of the basic single-pass filter Lman and the basic high-pass filter Hm sn, the passband of the frequency-gain characteristic is almost flat and the stopband The slope becomes gentle.
- a linear characteristic can be obtained in the frequency-phase characteristic. That is, it is possible to obtain a band-pass filter having good frequency characteristics without overshoot or ringing.
- n increases, the slope of the stop band becomes steeper, and the bandwidth of the pass band becomes narrower.
- the value of m is reduced, the slope of the stop band becomes steeper, and the bandwidth of the pass band becomes narrower.
- the numerical sequence has a contact point at a position where the original filter has a symmetrical first filter coefficient and a position where the frequency amplitude characteristic of the original filter takes a local maximum value.
- the filter design is performed by cascade-connecting an adjustment filter having a symmetrical second filter coefficient that realizes a frequency amplitude characteristic that has a minimum value at the contact point. Without using the equation, it is possible to easily design an FIR digital filter having an ideal frequency characteristic with a relatively wide passband and a steep stopband slope.
- the present invention provides an FIR digital filter of a type that includes a delay line with taps composed of a plurality of delay units, multiplies the output signal of each tap by a filter coefficient, adds the multiplication results thereof, and outputs the result. Useful.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Complex Calculations (AREA)
- Filters That Use Time-Delay Elements (AREA)
- Picture Signal Circuits (AREA)
- Tone Control, Compression And Expansion, Limiting Amplitude (AREA)
Abstract
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2005506147A JPWO2004102800A1 (ja) | 2003-05-15 | 2004-03-18 | デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム |
US10/556,775 US7590675B2 (en) | 2003-05-15 | 2004-03-18 | Digital filter, design method thereof, design device, and digital filter design program |
EP04721663A EP1626499A1 (en) | 2003-05-15 | 2004-03-18 | Digital filter, design method thereof, design device, and digital filter design program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2003-137134 | 2003-05-15 | ||
JP2003137134 | 2003-05-15 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2004102800A1 true WO2004102800A1 (ja) | 2004-11-25 |
Family
ID=33447247
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2004/003668 WO2004102800A1 (ja) | 2003-05-15 | 2004-03-18 | デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム |
Country Status (7)
Country | Link |
---|---|
US (1) | US7590675B2 (ja) |
EP (1) | EP1626499A1 (ja) |
JP (1) | JPWO2004102800A1 (ja) |
KR (1) | KR20060009238A (ja) |
CN (1) | CN1792033A (ja) |
TW (1) | TW200501565A (ja) |
WO (1) | WO2004102800A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2006048958A1 (ja) * | 2004-11-05 | 2006-05-11 | Neuro Solution Corp. | デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム |
US7590675B2 (en) | 2003-05-15 | 2009-09-15 | Nsc Co., Ltd. | Digital filter, design method thereof, design device, and digital filter design program |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2005104365A1 (ja) * | 2004-04-19 | 2005-11-03 | Neuro Solution Corp. | デジタルフィルタの設計方法および設計装置、デジタルフィルタ設計用プログラム、デジタルフィルタ、所望周波数特性の数値列の生成方法および生成装置、所望周波数特性の数値列生成用プログラム |
US8416861B2 (en) * | 2007-10-14 | 2013-04-09 | Nokia Corporation | Fixed-point implementation of an adaptive image filter with high coding efficiency |
CN101878648A (zh) * | 2007-10-14 | 2010-11-03 | 诺基亚公司 | 具有高编码效率的自适应图像滤波器的定点实现 |
JP5992634B2 (ja) * | 2013-10-31 | 2016-09-14 | 富士フイルム株式会社 | 信号処理装置、撮像装置、パラメータ生成方法、信号処理方法及びプログラム |
CN113965765A (zh) | 2018-03-09 | 2022-01-21 | 华为技术有限公司 | 使用自适应乘数系数进行图像滤波的方法及装置 |
EP3750307B1 (en) * | 2018-03-09 | 2023-10-04 | Huawei Technologies Co., Ltd. | Method and apparatus for image filtering with adaptive multiplier coefficients |
WO2021118427A1 (en) * | 2019-12-09 | 2021-06-17 | Telefonaktiebolaget Lm Ericsson (Publ) | Adaptive loop filtering |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002368584A (ja) * | 2001-06-06 | 2002-12-20 | Sony Corp | ディジタルフィルタおよびそれを用いたディジタルビデオエンコーダ |
WO2003023960A1 (fr) * | 2001-09-10 | 2003-03-20 | Neuro Solution Corp. | Filtre numerique et son procede de conception |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
SE517525C2 (sv) * | 1999-09-07 | 2002-06-18 | Ericsson Telefon Ab L M | Förfarande och anordning för konstruktion av digitala filter |
US6678708B1 (en) * | 2000-11-15 | 2004-01-13 | Intel Corporation | Method and apparatus for two-dimensional separable symmetric filtering |
US6993464B2 (en) * | 2002-12-03 | 2006-01-31 | Chunghwa Telecom Co., Ltd. | Optimized filter parameters design for digital IF programmable downconverter |
TW200501565A (en) | 2003-05-15 | 2005-01-01 | Neuro Solution Corp | Digital filter and design method, design apparatus, and digital filter design program thereof |
GB2427093A (en) * | 2004-02-17 | 2006-12-13 | Neuro Solution Corp | Digital filter design method and device, digital filter design program, and digital filter |
CN1965478A (zh) * | 2004-11-05 | 2007-05-16 | 神经网路处理有限公司 | 数字滤波器及其设计方法、设计装置、数字滤波器设计用程序 |
-
2004
- 2004-03-17 TW TW093107150A patent/TW200501565A/zh unknown
- 2004-03-18 CN CNA2004800133370A patent/CN1792033A/zh active Pending
- 2004-03-18 US US10/556,775 patent/US7590675B2/en not_active Expired - Fee Related
- 2004-03-18 WO PCT/JP2004/003668 patent/WO2004102800A1/ja active Application Filing
- 2004-03-18 KR KR1020057018016A patent/KR20060009238A/ko not_active Application Discontinuation
- 2004-03-18 EP EP04721663A patent/EP1626499A1/en not_active Withdrawn
- 2004-03-18 JP JP2005506147A patent/JPWO2004102800A1/ja active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002368584A (ja) * | 2001-06-06 | 2002-12-20 | Sony Corp | ディジタルフィルタおよびそれを用いたディジタルビデオエンコーダ |
WO2003023960A1 (fr) * | 2001-09-10 | 2003-03-20 | Neuro Solution Corp. | Filtre numerique et son procede de conception |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7590675B2 (en) | 2003-05-15 | 2009-09-15 | Nsc Co., Ltd. | Digital filter, design method thereof, design device, and digital filter design program |
WO2006048958A1 (ja) * | 2004-11-05 | 2006-05-11 | Neuro Solution Corp. | デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム |
Also Published As
Publication number | Publication date |
---|---|
US20060208831A1 (en) | 2006-09-21 |
CN1792033A (zh) | 2006-06-21 |
US7590675B2 (en) | 2009-09-15 |
EP1626499A1 (en) | 2006-02-15 |
KR20060009238A (ko) | 2006-01-31 |
TW200501565A (en) | 2005-01-01 |
JPWO2004102800A1 (ja) | 2006-07-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP1808962A1 (en) | Digital filter and its designing method, desiging apparatus, and program for designing digital filter | |
US7529788B2 (en) | Digital filter design method and device, digital filter design program, and digital filter | |
US11263293B2 (en) | Digital sample rate conversion | |
US20070053420A1 (en) | Method, apparatus, and program for designing digital filters | |
US7330865B2 (en) | Digital filter and its designing method | |
WO2004102800A1 (ja) | デジタルフィルタおよびその設計方法、設計装置、デジタルフィルタ設計用プログラム | |
EP1557947A1 (en) | Digital filter design method and device, digital filter design program, and digital filter | |
JPWO2004079905A1 (ja) | デジタルフィルタの設計方法および装置、デジタルフィルタ設計用プログラム、デジタルフィルタ | |
EP1533898A1 (en) | Digital filter designing method, digital filter designing program, digital filter | |
WO2003047097A1 (fr) | Procede de conception d'un filtre numerique, appareil de projet, programme de projet pour la realisation d'un tel filtre, et filtre numerique y relatif | |
JPWO2007102611A1 (ja) | 補間関数生成回路 | |
EP1557946A1 (en) | Digital filter design method and device, digital filter design program, and digital filter | |
US7400676B2 (en) | Tone quality adjustment device designing method and designing device, tone quality adjustment device designing program, and tone quality adjustment device | |
US20050171988A1 (en) | Digital filter design method and device, digital filter design program, and digital filter | |
JP2005020554A (ja) | デジタルフィルタ | |
US20050120067A1 (en) | Digital filter designing method, digital filter designing program, digital filter | |
Mottaghi-Kashtiban et al. | FIR filters involving shifts and only two additions, efficient for short word-length signal processing | |
JPH0346813A (ja) | デジタル・フィルタ回路 | |
Khorbotly et al. | Synthesis of recursive linear‐phase filters for fixed‐point hardware platforms | |
JPWO2003096534A1 (ja) | 音質調整装置の設計方法および設計装置、音質調整装置設計用プログラム、音質調整装置 | |
JP2012085177A (ja) | デシメータ回路及びデシメータ回路の演算方法 | |
JP2001339277A (ja) | ディジタルfirフィルタ | |
JPH10150344A (ja) | 再帰型デジタルフィルタ |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): BW GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 1020057018016 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2005506147 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2004721663 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 20048133370 Country of ref document: CN Ref document number: 10556775 Country of ref document: US |
|
WWP | Wipo information: published in national office |
Ref document number: 1020057018016 Country of ref document: KR |
|
WWP | Wipo information: published in national office |
Ref document number: 2004721663 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 10556775 Country of ref document: US |