Innovation Card

Andrew Maffei (WHOI) edited this page Jan 17, 2019 · 12 revisions

Innovation Card: GobySoft Progressive Imagery Payload

Introduction

This payload is a hybrid hardware/software rapid prototyping effort as a collaboration between GobySoft and the WHOI Innovation Project for use on a variety of mooring and AUV systems. The hardware is designed from low-cost IoT components and built using Gumstix's Geppetto service, and runs a Raspberry Pi Compute Compute module and various communications interfaces.

This payload was deployed in the Imaging Flow CytoBot (IFCB) with software that is a blend of open source projects to demonstrate progressive image transmission over slow links: the Goby3 middleware from GobySoft, the Progressive Imagery code from WHOI NDSF, and cgsn-mooring software from WHOI OOI. This software suite allows online transmission of IFCB images with improving quality as more network throughput is available. In addition, we have a placeholder for classification using future advanced techniques such as machine learning algorithms.

Services Provided

  • Progressive transmission over slow links of visual imagery. The operator receives a progressively enhance image as more data are available shore side:
    • Example of single progressive image (on loop):
    • Prototype user interface for multiple images:
  • Stub for classification of images
    • Only images exceeding a certain size and of a defined aspect ratio are transmitted.
    • This stub wip_comms_classifier could be expanded to take advantage of machine learning techniques.

Connections

Physical

  • Payload PCB designed to be integrated and mounted in pressure housing.
  • Raspberry Pi Compute Module 3 (CM3) uses standard SO-DIMM RAM physical connector to attach to Payload board
  • USB connectors use standard ATX motherboard 9-pin dual USB headers

Power

  • Power connector is 1.7mm ID, 4.0mm OD barrel plug
  • Nominal Power (under benchtest load for 2018 Scripps IFCB deployment)
    • Voltage: 5V
    • Power: 2.5W
  • Under Load Power (using stress -c 4 CPU stress test)
    • Voltage: 5V
    • Power: 5W

Software / Communications

IFCB Deployment Software Structure Diagram

Parts

Hardware

Software

  • Demo Software:
    • Description: wip-comms-demo:* Libraries and Goby3 Applications for this demonstration
    • Owner: WHOI Innovation Project
    • Source code
    • Documentation
  • Communications Middleware:
  • Comms Compression Software:
  • Image Delivery Software:
    • Description: Progressive Imagery library and ROS/Goby3 applications
      • Owner: WHOI NDSF
      • License: Not currently open source; hopefully in the near future.
  • Platform Control Software:
  • Instrument/Sensor Software:
    • Description: pyifcb - Imaging FlowCytobot Python API, generation 2
    • Description: wip-comms-ifcb-imagedb: Writes IFCB images and minimal metadata to a PostgreSQL database in ppm format
Clone this wiki locally
You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.
Press h to open a hovercard with more details.