OpenOceanCam, 2019

Open Ocean Camera Project

The Open Ocean Camera

Our goal is to leverage the power of technology and imaging to make the marine conservation process affordable, efficient and scalable.

Current marine monitoring with the GoPro is expensive and impractical. We're building a camera suited to the needs of marine scientists. 

  • The camera comes with 1Tb SSD memory and 3 weeks of battery life. This eliminates the need to rent a boat and dive every three days saving you hundreds of dollars over a year!

  • The camera shooting timetable can be set for a few weeks in advance. Over 25 shooting modes supported. Optimize your photographs to your need and save resources at night.

  • For each time-slot the ISO, Shutter-Speed, Focal Length, Interval can be customized. Video Mode is also supported.

  • The camera has an on-board raspberry pi Linux computer allowing you to run biomass estimation and other algorithms in real-time!

  • We are building a GUI deep learning interface so that you can train and test the machine learning kernel to auto-classify animals without typing a single line of code.

  • The camera can auto-classify images and saving only the frames where an animal is detected, saving you hours of manual classification work.

IMG_20190303_180540.jpg
The App
Easy Deep Learning (Ongoing Completion)
  • Select the folder with your images

  • Select an excel file with the image names and their classification

  • Click the "train" button

  • Select a test image

  • Click the "test" button 

  • View Result

  • Connect to the camera through Wi-Fi

  • Set the time slot parameters

  • Send to the camera

  • Edit/View the configuration as required

Progress & Open Source Release

We've been working with National Geographic Explorer Dr. Alessandro Ponzo and the Large Marine Vertebrates Research Institute Philippines to build this camera.

January 2019: Project commencement 

February 2019: Round 1 Prototyping in Hong Kong

March 2019: Expedition to El Nido and Testing

 

Manta photographed by the OpenOceanCam in Sibiltan, Philippines

 

DSC_1400.JPG
Team and Sponsors

OpenOceanCam was started at the University of Hong Kong by Sidhant and Saumya Gupta with the help of Dr. Alessandro Ponzo of the Large Marine Vertebrates Research Institute Philippines. 

The project was awarded a travel grant by the Gallant Ho Experiential Learning Fund at the University of Hong Kong.

The March 2019 Sibilitan expedition team:

  1. Sidhant Gupta - Founder, project lead (India)

  2. Saumya Gupta - Founder, co-project lead (India)

  3. Dominic Co - Design (Philippines, Canada)

  4. Braedyn Au - Coding (Canada)

  5. Andy Vu - Biology (Portuguese, Macau SAR)

  6. Kengie Tang - Biology & Documentation (Singapore)

  7. Camen Wong - Diving Team (Hong Kong SAR)

  8. Rachael Chu - Diving Team (Hong Kong SAR)

  9. Sovina Taneja - Team Admin and Expedition Logistics (India)

  10. Utkarsh Goyal - Computer Engineering (India)

  11. Winnie Suen - Documentation (Hong Kong SAR)

  12. Iqra Abbas - Design and Coding (Pakistan)

  13. Rajat Jain - Coding (India)

  14. Maria Kiejnich - Project Dissemination & Impact Scaling (Poland)

We'd also like to extend our thanks to Dr. Alessandro Ponzo of the LAMAVE institute for his extended support of this project and to Mo, Sana and Mona at the Dive Shop Sibiltan for making our expedition possible. 

Contact Us

OpenOceanCam

1011B, Lee Shau Kee Hall, JCSV2,

109, Pokfulam Road, Hong Kong

(+852) 51385925

sdhnt@connect.hku.hk