|
DETAILS
|
Sum
|
Sum
FILTER
|
| |
| |
Sums up sample values along dimensions. |
| |
| |
+ |
mean
|
BOOL
|
1
|
'false'
|
LOCK
|
replace sum by mean value
|
| |
|
Statistics
|
Statistics
FEATURE
|
| |
| |
Raises statistical evaluations on the input stream. |
| |
| |
+ |
A-kurtosis
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs kurtosis as a dimension
|
| |
+ |
B-skewness
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs skewness as a dimension
|
| |
+ |
C-mean
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs mean as a dimension
|
| |
+ |
D-stddev
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs stddev as a dimension
|
| |
+ |
E-var
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs var as a dimension
|
| |
+ |
F-number-of-values
|
BOOL
|
1
|
'false'
|
LOCK
|
outputs the number of values as a dimension
|
| |
|
Spectrogram
|
Spectrogram
FEATURE
|
| |
| |
Computes spectrogram of input stream. |
| |
| |
+ |
nfft
|
UINT
|
1
|
'512'
|
LOCK
|
#fft coefficients
|
| |
+ |
nbanks
|
UINT
|
1
|
'50'
|
LOCK
|
#filter banks
|
| |
+ |
minfreq
|
DOUBLE
|
1
|
'0.00000'
|
LOCK
|
mininmum frequency
|
| |
+ |
maxfreq
|
DOUBLE
|
1
|
'0.00000'
|
LOCK
|
maximum frequency (nyquist if 0)
|
| |
+ |
wintype
|
INT
|
1
|
'3'
|
LOCK
|
window type (0=rectangle,1=triangle,2=gauss,3=hamming
|
| |
+ |
dolog
|
BOOL
|
1
|
'true'
|
LOCK
|
apply logarithm
|
| |
+ |
dopower
|
BOOL
|
1
|
'true'
|
LOCK
|
compute the PSD for every bank
|
| |
+ |
file
|
CHAR
|
1024
|
''
|
LOCK
|
file with filter banks (plain text file with each line defining one band { })
|
| |
+ |
banks
|
CHAR
|
1024
|
''
|
LOCK
|
string with filter banks that gets applied if no file was set (example: "0.003 0.040
0.040 0.150
0.150 0.400").
|
| |
|
Relative
|
Relative
FILTER
|
| |
| |
Makes each value of each dimension relative to the value before. |
| |
| |
|
Pulse
|
Pulse
FEATURE
|
| |
| |
Computes pulse (in Hz) based on distance of successive maxima. |
| |
| |
+ |
delta
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
minimum difference in amplitude between successive maxima (to filter local maxima)
|
| |
+ |
minr
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
minimum rate in hz
|
| |
+ |
maxr
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
maximum rate in hz
|
| |
|
Normalize
|
Normalize
FILTER
|
| |
| |
Normalizes input stream into fixed interval. |
| |
| |
+ |
min
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
minimum value
|
| |
+ |
max
|
FLOAT
|
1
|
'1.00000'
|
LOCK
|
maximum value
|
| |
|
Noise
|
Noise
FILTER
|
| |
| |
Adds normally distributed noise in certain frequency range to input stream. |
| |
| |
+ |
ampl
|
FLOAT
|
1
|
'1.00000'
|
LOCK
|
noise amplitude
|
| |
+ |
mean
|
DOUBLE
|
1
|
'0.00000'
|
LOCK
|
noise mean
|
| |
+ |
stdv
|
DOUBLE
|
1
|
'1.00000'
|
LOCK
|
noise standard deviation
|
| |
+ |
norm
|
BOOL
|
1
|
'true'
|
LOCK
|
frequency values are normalized in interval [0..1], where 1 is the nyquist frequency (=half the sample rate)
|
| |
+ |
cutoff
|
DOUBLE
|
1
|
'0.00000'
|
LOCK
|
if > 0: cutoff frequency either in hz or normalized (see -norm)
|
| |
+ |
width
|
DOUBLE
|
1
|
'0.00000'
|
LOCK
|
if > 0: band width either in hz or normalized (see -norm)
|
| |
|
MvgPeakGate
|
MvgPeakGate
FILTER
|
| |
| |
Blocks out everything but peaks in the input stream. |
| |
| |
+ |
value
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
blocked samples are set to this value
|
| |
+ |
keep
|
BOOL
|
1
|
'false'
|
LOCK
|
for unblocked samples keep original values
|
| |
+ |
thres
|
INT
|
1
|
'0'
|
LOCK
|
threshold (0=fix,1=fix+avg,2=fix+avg+mstd*std)
|
| |
+ |
fix
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
value of fixed threshold
|
| |
+ |
mstd
|
FLOAT
|
1
|
'1.00000'
|
LOCK
|
multiplier for standard deviation adjusted threshold
|
| |
+ |
full
|
BOOL
|
1
|
'false'
|
LOCK
|
use all values to update threshold, otherwise only blocked samples are used
|
| |
+ |
win
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
+ |
monitor
|
BOOL
|
1
|
'false'
|
LOCK
|
display threshold monitor
|
| |
+ |
mpos
|
INT
|
4
|
'0,0,280,80'
|
LOCK
|
position of monitor on screen [posx,posy,width,height]
|
| |
|
MvgNorm
|
MvgNorm
FILTER
|
| |
| |
Normalizes input stream using moving/sliding minim and/or maximum for a chosen window. |
| |
| |
+ |
norm
|
INT
|
1
|
'0'
|
LOCK
|
norm (0=meanvar,1=minmax,2=subavg,3=submin)
|
| |
+ |
rangea
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
lower interval border
|
| |
+ |
rangeb
|
FLOAT
|
1
|
'1.00000'
|
LOCK
|
uper interval border
|
| |
+ |
win
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
+ |
nblock
|
UINT
|
1
|
'10'
|
LOCK
|
n blocks (for moving filter only)
|
| |
|
MvgMinMax
|
MvgMinMax
FILTER
|
| |
| |
Computes moving/sliding minim and/or maximum of the input stream for a chosen window. |
| |
| |
+ |
win
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
+ |
format
|
INT
|
1
|
'3'
|
LOCK
|
format (1=min,2=max,3=both)
|
| |
+ |
nblock
|
UINT
|
1
|
'10'
|
LOCK
|
n blocks (for moving filter only)
|
| |
|
MvgMedian
|
MvgMedian
FILTER
|
| |
| |
Computes median of the input stream for a chosen window. |
| |
| |
+ |
win
|
DOUBLE
|
1
|
'1.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
winInSamples
|
UINT
|
1
|
'0'
|
LOCK
|
size of moving/sliding window in samples (overrides win)
|
| |
|
MvgDrvtv
|
MvgDrvtv
FILTER
|
| |
| |
Computes moving/sliding derivative of the input stream for a chosen window. |
| |
| |
+ |
win
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
|
MvgConDiv
|
MvgConDiv
FILTER
|
| |
| |
Computs moving/sliding convergence/divergence for a chosen window. |
| |
| |
+ |
wins
|
DOUBLE
|
1
|
'5.00000'
|
LOCK
|
size of short moving/sliding window in seconds
|
| |
+ |
winl
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of long moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
|
MvgAvgVar
|
MvgAvgVar
FILTER
|
| |
| |
Computes moving/sliding average and/or variance of the input stream for a chosen window |
| |
| |
+ |
win
|
DOUBLE
|
1
|
'10.00000'
|
LOCK
|
size of moving/sliding window in seconds
|
| |
+ |
method
|
INT
|
1
|
'0'
|
LOCK
|
method (0=moving,1=sliding)
|
| |
+ |
format
|
INT
|
1
|
'3'
|
LOCK
|
format (1=average,2=variance,3=both)
|
| |
|
Multiply
|
Multiply
FILTER
|
| |
| |
Multiplies input stream with a constant factor. |
| |
| |
+ |
single
|
BOOL
|
1
|
'true'
|
LOCK
|
a single multiplication factor is given (factor), otherwise one per dimension (factors)
|
| |
+ |
factor
|
FLOAT
|
1
|
'1.00000'
|
FREE
|
multiplication factor
|
| |
+ |
factors
|
CHAR
|
1024
|
''
|
FREE
|
multiplication factors separated by comma
|
| |
+ |
join
|
INT
|
1
|
'0'
|
FREE
|
join dimensions (0=off, 1=multiply, 2=sum up, 3=sum up squares
|
| |
|
Mean
|
Mean
FEATURE
|
| |
| |
Mean values along dimensions. |
| |
| |
|
MFCC
|
MFCC
FEATURE
|
| |
| |
Computes mel frequency cepstral coefficients (MFCCs) of input stream. |
| |
| |
+ |
n_first
|
UCHAR
|
1
|
'0'
|
LOCK
|
first cepstral
|
| |
+ |
n_last
|
UCHAR
|
1
|
'13'
|
LOCK
|
last cepstral
|
| |
+ |
n_ffts
|
UCHAR
|
1
|
'0'
|
LOCK
|
fft size
|
| |
|
Limits
|
Limits
FILTER
|
| |
| |
Limits lower and upper range of a signal to the given values. |
| |
| |
+ |
min
|
FLOAT
|
1
|
'0.00000'
|
FREE
|
minimum value (lower range)
|
| |
+ |
max
|
FLOAT
|
1
|
'1.00000'
|
FREE
|
maximum value (upper range)
|
| |
+ |
replaceZero
|
BOOL
|
1
|
'false'
|
FREE
|
replace data out of range with zero (false by default)
|
| |
|
Intensity
|
Intensity
FEATURE
|
| |
| |
Computes intensity of input stream. |
| |
| |
+ |
global
|
BOOL
|
1
|
'false'
|
LOCK
|
calculate energy over dimensions
|
| |
|
Integral
|
Integral
FILTER
|
| |
| |
Computes integral of input stream by summing consecutive samples. |
| |
| |
+ |
names
|
CHAR
|
1024
|
''
|
LOCK
|
names of integral separated by blank (0th,1st,2nd,3rd,4th) or leave empty to select all)
|
| |
+ |
reset
|
BOOL
|
1
|
'false'
|
LOCK
|
always reset
|
| |
|
IIR
|
IIR
FILTER
|
| |
| |
Infinite impulse response (IIR) filter. |
| |
| |
|
Gate
|
Gate
FILTER
|
| |
| |
Replaces values below threshold. |
| |
| |
+ |
thres
|
FLOAT
|
1
|
'0.00000'
|
FREE
|
threshold (everything below is set to 'replace')
|
| |
+ |
replace
|
FLOAT
|
1
|
'0.00000'
|
FREE
|
replace value
|
| |
+ |
upper
|
BOOL
|
1
|
'false'
|
FREE
|
applies threshold as upper limit
|
| |
|
FunctionalsEventSender
|
FunctionalsEventSender
CONSUMER
|
| |
| |
Applies functionals to the current stream and sends the result as an event. |
| |
| |
+ |
address
|
CHAR
|
1024
|
''
|
LOCK
|
event address (event@sender)
|
| |
+ |
sname
|
CHAR
|
1024
|
'floats'
|
LOCK
|
name of sender [deprecated, see address]
|
| |
+ |
ename
|
CHAR
|
1024
|
'floats'
|
LOCK
|
name of event [deprecated, see address]
|
| |
+ |
names
|
CHAR
|
1024
|
''
|
LOCK
|
names of functionals separated by blank (mean,energy,std,min,max,range,minpos,maxpos,zeros,peaks,len) or leave empty to select all)
|
| |
+ |
delta
|
UCHAR
|
1
|
'2'
|
LOCK
|
zero/peaks search offset
|
| |
|
Functionals
|
Functionals
FEATURE
|
| |
| |
Computes a series of functionals from input stream. |
| |
| |
+ |
names
|
CHAR
|
1024
|
''
|
LOCK
|
names of functionals separated by comma (mean,energy,std,min,max,range,minpos,maxpos,zeros,peaks,len,path) or leave empty to select all except path
|
| |
+ |
delta
|
UCHAR
|
1
|
'2'
|
LOCK
|
zero/peaks search offset
|
| |
+ |
global
|
BOOL
|
1
|
'false'
|
LOCK
|
calculate over all dimensions
|
| |
|
FFTfeat
|
FFTfeat
FEATURE
|
| |
| |
Applies fast fourier transformation and outputs coefficients as feature values. |
| |
| |
+ |
nfft
|
UINT
|
1
|
'512'
|
LOCK
|
#fft coefficients
|
| |
+ |
wintype
|
INT
|
1
|
'3'
|
LOCK
|
window type (0=rectangle,1=triangle,2=gauss,3=hamming
|
| |
|
Expression
|
Expression
FILTER
|
| |
| |
Applies mathematical expressions to a stream (see http://www.speqmath.com/tutorials/expression_parser_cpp/index.html for supported operators). |
| |
| |
+ |
single
|
BOOL
|
1
|
'false'
|
LOCK
|
apply a single expression to all dimensions otherwise expression is applied to each dimension (use d0, d1, ... to address specific dimension or just d otherwise)
|
| |
+ |
expression
|
CHAR
|
1024
|
'd + d'
|
LOCK
|
expression string (e.g. 'd+d' doubles each dimension, 'd0+d1+d2..' sums up dimensions)
|
| |
+ |
join
|
INT
|
1
|
'0'
|
LOCK
|
only applies if single==false : join dimensions after applying expression (0=off, 1=multiply, 2=sum up
|
| |
+ |
cast
|
UINT
|
1
|
'9'
|
LOCK
|
cast result after expression was applied (CHAR = 1, UCHAR = 2, SHORT = 3, USHORT = 4, INT = 5, UINT = 6, LONG = 7, ULONG = 8, FLOAT = 9, DOUBLE = 10, LDOUBLE = 11, STRUCT = 12, IMAGE = 13, BOOL = 14
|
| |
|
Energy
|
Energy
FEATURE
|
| |
| |
Computes energy of input stream. |
| |
| |
+ |
global
|
BOOL
|
1
|
'false'
|
LOCK
|
calculate energy over dimensions
|
| |
|
DownSample
|
DownSample
TRANSFORMER
|
| |
| |
Downsamples a signal keeping only every x'th sample. |
| |
| |
+ |
keep
|
UINT
|
1
|
'1'
|
LOCK
|
keep every n'th sample (> 0)
|
| |
+ |
remove
|
BOOL
|
1
|
'false'
|
LOCK
|
instead of keeping remove samples
|
| |
+ |
mean
|
BOOL
|
1
|
'false'
|
LOCK
|
output mean of last n'th values (only applied if remove=false)
|
| |
|
Derivative
|
Derivative
FILTER
|
| |
| |
Computes derivative of input stream by subtracting consecutive samples. |
| |
| |
+ |
names
|
CHAR
|
1024
|
''
|
LOCK
|
names of derivatives separated by blank (0th,1st,2nd,3rd,4th) or leave empty to select all)
|
| |
|
ConvPower
|
ConvPower
FEATURE
|
| |
| |
Computes energy from the convolution of the input stream. |
| |
| |
+ |
global
|
BOOL
|
1
|
'false'
|
LOCK
|
calculate energy accross all dimensions
|
| |
|
Butfilt
|
Butfilt
FILTER
|
| |
| |
Applies butter low/high/bandpass filter to input stream. |
| |
| |
+ |
type
|
UCHAR
|
1
|
'0'
|
LOCK
|
filter type (0=low, 1=high, 2=band)
|
| |
+ |
order
|
UINT
|
1
|
'1'
|
LOCK
|
filter order
|
| |
+ |
norm
|
BOOL
|
1
|
'true'
|
LOCK
|
frequency values are normalized in interval [0..1], where 1 is the nyquist frequency (=half the sample rate)
|
| |
+ |
low
|
FLOAT
|
1
|
'0.00000'
|
FREE
|
low cutoff frequency given either as normalized value in interval [0..1] or as an absolute value in Hz (see -norm)
|
| |
+ |
high
|
FLOAT
|
1
|
'1.00000'
|
FREE
|
high cutoff frequency given either as normalized value in interval [0..1] or as an absolute value in Hz (see -norm)
|
| |
+ |
zero
|
BOOL
|
1
|
'true'
|
LOCK
|
subtract first sample from signal to avoid artefacts at the beginning of the signal
|
| |
|
Bundle
|
Bundle
FILTER
|
| |
| |
Removes outliers above or below a threshold. |
| |
| |
+ |
thres
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
threshold to find outliers
|
| |
+ |
hang_in
|
UINT
|
1
|
'0'
|
LOCK
|
#successive values above threshold to change direction
|
| |
+ |
hang_out
|
UINT
|
1
|
'0'
|
LOCK
|
#successive values below threshold to change direction
|
| |
+ |
replace_below
|
INT
|
1
|
'0'
|
LOCK
|
method to replace outliers below threshold: 0=PREVIOUS, 1=EXTREMA, 2=MEAN, 3=THRESHOLD, 4=SET
|
| |
+ |
replace_above
|
INT
|
1
|
'0'
|
LOCK
|
method to replace outliers above threshold: 0=PREVIOUS, 1=EXTREMA, 2=MEAN, 3=THRESHOLD, 4=SET
|
| |
+ |
set_below
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
if 'replace_below=SET' this value will be used
|
| |
+ |
set_above
|
FLOAT
|
1
|
'0.00000'
|
LOCK
|
if 'replace_above=SET' this value will be used
|
| |
+ |
set_all
|
BOOL
|
1
|
'false'
|
LOCK
|
set all values to either 'set_below' or 'set_above'
|
| |
|