Evan Mikel Lapisky

US Patent:

20050007880, Jan 13, 2005

Filed:

Jun 1, 2004

Appl. No.:

10/856871

Inventors:

Matthew Zimmerman - Providence RI, US
Matthew Coolidge - Narragansett RI, US
Evan Lapisky - Narragansett RI, US

International Classification:

G01S015/00

US Classification:

367103000

Abstract:

A processing technique and transducer field of view architecture for use as part of an imaging sonar system to develop three-dimensional images of the space below, to the sides, and forward of a ship.

US Patent:

20070159922, Jul 12, 2007

Filed:

Oct 16, 2006

Appl. No.:

11/581626

Inventors:

Matthew Zimmerman - Providence RI, US
Matthew Coolidge - Providence RI, US
Evan Lapisky - Narragansett RI, US

International Classification:

G01S 15/00

US Classification:

367103000

Abstract:

A 3-dimensional sonar system having a fixed frame of reference including a transmitter and receiver array. The system may include a single ping processor for processing received echoes from a single transmission including methods to match filter sonar sensor data, to optionally compensate for self Doppler, beamform sonar sensor data, to extract bottom targets from beamformed data and to extract in-water targets from the beamformed data. The system may also include a multi ping processor to operate on the outputs of multiple pings from the single ping processor including methods to detect tracks from in-water targets.

US Patent:

20200379104, Dec 3, 2020

Filed:

Mar 5, 2020

Appl. No.:

16/809983

Inventors:

- Warwick RI, US
Heath Henley - Warwick RI, US
Austin Berard - North Smithfield RI, US
Evan Lapisky - South Kingstown RI, US

International Classification:

G01S 15/04
G01S 7/52
G01S 15/89
G01S 7/53

Abstract:

Machine learning algorithms can interpret three-dimensional sonar data to provide more precise and accurate determination of seafloor depths and in-water target detection and classification. The models apply architectures for interpreting volumetric data to three-dimensional forward-looking sonar data. A baseline set of training data is generated using traditional image and signal processing techniques, and used to train and evaluate a machine learning model, which is further improved by additional inputs to improve both seafloor and in-water target detection.