Attenuated Signal Removal

This operator identifies and adjusts pings which show an attenuated signal when compared to surrounding pings. Attenuated signal can arise from air bubble attenuation within the water column below the transducer.

The operator is based on the “attenuated signal (AS) filter” algorithm and definitions described in Ryan et al. (2015).

It accepts operands with the following data types:

Complex power dB
Complex Sv
Complex TS
Power dB
Pulse compressed complex power dB
Pulse compressed complex Sv
Pulse compressed complex TS
Sv
TS
Unspecified dB

Settings

The Attenuated Signal Removal Variable Properties dialog box pages include (common) Variable Properties pages and these operator pages:

Operands page

Attenuated Signal Removal page

Detection

Setting	Description
Exclude above	Specifies the exclude above line. Ryan et al observed that the attenuated signal algorithm is optimized when working with a deep scattering layer where signal to noise is high and ping overlap is high (assuming no vessel rotation). Generally, such a layer provides a relatively homogeneous region in which to assess pings with attenuated signal. In the case where a deep scattering layer is weak or non-existent, the sea floor region may provide a useful reference layer. The use of exclusion lines can improve operator performance.
Exclude below	Specifies the exclude below line. Ryan et al observed that the attenuated signal algorithm is optimized when working with a deep scattering layer where signal to noise is high and ping overlap is high (assuming no vessel rotation). Generally, such a layer provides a relatively homogeneous region in which to assess pings with attenuated signal. In the case where a deep scattering layer is weak or non-existent, the sea floor region may provide a useful reference layer. The use of exclusion lines can improve operator performance.
Context window size (pings)	Specifies the extent of the surrounding pings on either side of a center ping. Attenuated pings can occur as single pings or a group of successive pings. Therefore, the value of Context window size affects the detection of attenuated signal pings and the replacement of samples in identified pings. Ryan et al have found that the value of Context window size 'is a compromise and is chosen based on an inspection of the echogram to match the typical duration of contiguous attenuated pings.' If the size is too small the percentile of samples in the center ping becomes too similar to the percentile of the samples in surrounding pings. If the size is too large and the signal in the deep scattering layer is too variable or too weak, the window sample percentile may not be a useful local reference value. Echoview requires odd values for the Context window size and will round up for even numbers. The default value for Context window size is 31 pings.
Percentile	Percentile specifies the value for the Nth Percentile algorithm for the context pings.
Threshold (dB)	Specifies the attenuated signal threshold. The current ping is deemed to be an attenuated signal when the difference between the percentile score of samples in the surrounding pings and the percentile score of samples in the current ping is greater than the Threshold.

Replacement

Setting

Description

Replacement value

Specifies the replacement method used for attenuated signal samples.

No data
The ping is replaced by no data.

Percentile

The ping is replaced by the percentile score of samples at the same range in neighboring pings.

Percentile

Percentile specifies the value for the Nth Percentile algorithm for the Percentile Replacement method.

Attenuated signal removal algorithm

This algorithm identifies and adjusts sample values that are significantly lower than those of surrounding samples.

You can specify settings on the Attenuated Signal Removal page of the Variable Properties dialog box.

Notes:

No data samples within the Context Window are not used in percentile calculations and are not replaced.
Context window size affects the detection and replacement of attenuated pings. The larger the Context window size the longer the processing time. Specifying exclusion lines that delineate a deep scattering layer can improve the detection of attenuated signal pings and minimize the processing time.
The operator algorithm may fail to identify attenuated pings in the set of start and end pings of the echogram. This is due to the handling of boundary samples with respect to the Context window. To address this issue, you can inspect the set of first and last pings and use regions and/or advanced operators to deal with manually identified attenuated signal pings.
Samples in adjacent pings may shift in depth under heave compensation and this may affect the attenuated signal pings.
Under the Replacement step, adjacent samples that are not at the same depth, retain their original value.
Complex and pulse compressed complex acoustic variables are largely handled in a manner similar to single frequency power, Sv or TS variables. When a new sample value is required, it is assumed that the complex phase for the sample is unchanged and the complex magnitude of the sample is calculated as for ComplexMagnitudeNew.
See also: The Data thresholding section on the Noise, background object and signal removal in Echoview page.
See also: The Effect of No data samples on the Wideband Frequency Response graph page.

Context window for attenuated signal detection and sample replacement

The algorithm and variable properies dialog box setting described in this section is shared between the Attenuated Signal Removal (this page) and the Transient noise ping removal operator. We indicate any differences within the text.

Attenuated signal window

A Context window consists of a center ping and surrounding pings. Samples in identified pings are replaced using a specified Replacement value.

Calculate the percentile score for the samples in the center ping of the Context Window (between Exclude lines if specified).
Calculate the percentile score for the samples of the Context window pings surrounding the ping of interest.
If the central ping's percentile score is lower than the Context window percentile score by more than a specified Threshold then consider the center ping to be attenuated.
Important: For the Transient noise ping removal operator, this is |vij - vmn| > ThresholdAmplified signal instead.
Samples in an attenuated ping are replaced by the specified Replacement value.
1. When the Replacement value is No data, all samples in the center ping are set to a no data.
2. When the Replacement value is Percentile, the sample in the center ping is set to the percentile score of the horizontal surrounding samples in the Context window.

Context Window percentiles

The Context Window Percentile Lower percentile of samples in the context window represents the percentile score of the sample values surrounding the center ping.

Let vp be the value of the Context Window Pth-Percentile of an ascending order data set containing N elements with values. The N elements are the samples values of the surrounding pings within the Context Window, and exclude center ping sample values. The element values are ordered as follows:

Calculate:

Where P is specified by Percentile on the Attenuated Signal Removal/Transient Noise Ping Removal page of the virtual variable.

The center ping percentile vij represents the percentile score of samples in the center ping of the Context Window.

Replacement value percentile

The Replacement value Percentile vp is used to replace a center ping attenuated sample value.

Let vp be the value of the Sample Pth-Percentile of an ascending order data set containing N elements with values. The N elements are the horizontal samples values on either side of the center sample within the Context Window. The element values are ordered as follows:

Calculate:

Where P is specified by Replacement Percentile on the Attenuated Signal Removal/Transient Noise Ping Removal page of the virtual variable.

Percentile algorithm

Where:

k'
=

Percentile rank where:

P is the Pth-percentile.

N is the number of elements in the ascending order set of values.

k

=
where Floor function , that gives the largest integer less than or equal to k'.