Description
The ClassifyHIPSNoise process is used to automatically identify and remove noise from bathymetric data, using a pre-trained Convolutional Neural Network (CARIS Mira AI) to assign a Noise Confidence value to each point. The Noise Confidence is a percentage between 0 and 100%, 100% meaning 100% confidence that the point is noise.
A filtering process can be applied during or after the noise classification process, where any points with a Noise Confidence above the user-defined threshold are rejected. The default threshold is 50%.
Update
The source point attributes are updated as appropriate by the selected algorithm in the HIPS track lines.
Command line syntax
carisbatch --run ClassifyHIPSNoise [options] <update>
User Interface Command
Sonar Noise Classifier
Options
The following table describes the parameters available for this process.
Parameter | ||||
|---|---|---|---|---|
Long | Short key | Description | Can Repeat | Notes
|
--mira-url | The URL of the Mira service. The default is http://username:password@carismira.ai username and password must be replaced with the credentials for an account that has access to CARIS Mira. If credentials are not provided, the application will check the Windows Credential Manager for credentials that have been saved for the default URL. If not found, the process will fail. | FALSE | ||
--finest-vertical-resolution | If set, a number specifying the finest vertical spacing allowed between sampling bins to be sent to the classifier. The default is 0.1m. A lower threshold means greater vertical exaggeration is allowed, and more precise classification of points. However, this can result in the removal of real features when there is a short, abrupt change in slope. The maximum number of vertical bins allowed is 100, which will override this setting if needed. Supported units: cable, ch, cm, fm, ft, in, inm, km, m, mi, mm, nm, usfm, usft, usmi, usnm, usyd, yd. | FALSE | ||
--noise-confidence-filter-threshold | If set, the noise confidence filter threshold, between 0 and 100, will mark as Rejected (Noise Confidence/Polygon Filter) any classified points with a confidence above this threshold. | FALSE | ||
--level-of-detail | The level of detail sent to the noise classifier. • Fine • Medium • Coarse The default is Medium. A finer level of detail means a more precise classification, but at the expense of more requests to the AI and longer processing times. | FALSE | ||
Example
To use this command refer to the following example.
Example | ||
|---|---|---|
Objective | You want to use the noise classifier to remove noise from your sonar data. | |
Description | Command Line Syntax | |
The following credentials will be used to connect to CARIS Mira. • Username: John Doe • Password: Default! | --mira-url http://johndoe:default!@carismira.ai | |
Classification will be performed using the finest level of detail. | --level-of-detail FINE | |
All points with a classification confidence greater than 75% will be flagged as Rejected. | --noise-confidence-filter-threshold 75 | |
The sonar data is called Demo.hips and can be found in D:/Data/Demo. | D:/Data/Demo/Demo.hips | |
Command Line | carisbatch -r ClassifyHIPSNoise --mira-url http://johndoe:default!@carismira.ai --level-of-detail FINE --noise-confidence-filter-threshold 75 D:/Data/Demo/Demo.hips | |
Outcome | All points classified as noise in Demo.hips are flagged as rejected. | |
Extra Notes | ||