An AI filter that significantly enhances underwater visuals
An AI filter that significantly enhances underwater visuals.
The need for frequent inspection and maintenance is greater than ever with the increase in offshore oil and gas facilities. Underwater robots, whether remotely operated underwater vehicles (ROVs) or autonomous underwater vehicles (AUVs), have proven to be an asset for offshore operators in these areas, creating a cost-efficient solution that cuts out the need for humans to dive down to the seabed.
However, to inspect the structures placed on the seabed effectively, manufacturers must equip ROVs and AUVs with cameras that provide the best possible visuals. The better the visuals, the more informed operators will be when making decisions. This is something Chinese robotics company Qysea Technology had in mind when developing its FiFish E-Go underwater robot.
Developers at Qysea equipped the robot with an industry-first 176° ultra-wide fisheye lens, delivering a 146° field of vision underwater to capture a wide range of elements in a single frame.
With a macro adjustment capability, the FiFish E-Go focuses underwater at a 10cm distance, allowing for clear observation at extremely close distances, according to the company. It also features 10,000 lumens of underwater lighting and a 160° wide illumination angle, ensuring clarity under dark and turbid oceans.
AI-DRIVEN UNDERWATER OPERATIONS
Qysea’s self-developed AI Vision Lock function enables FiFish to achieve adaptive stabilisation and locking of target objects underwater with a simple touch on a control screen. Furthermore, its AI plankton filtering algorithm automatically optimises visuals, ensuring clearer underwater scenes for professional underwater monitoring tasks.
Its AI Dehazing Algorithm automatically identifies and filters out the snowflake effect caused by underwater suspended particles, enhancing image clarity, and improving decision-making efficiency for underwater operations.
“A clearer vision underwater offers significant advantages for [oil and gas] operators,” the Qysea technical team tells IME. “This includes enhanced visual data for detailed assessments, optimised time and cost efficiency of inspections, as well as increased diver safety when monitoring underwater maintenance operations.”
UNDERWATER ALGORITHMS
The Qysea technical team explain that techniques such as dehazing, contrast enhancement, and natural color correction, mean these algorithms optimise underwater images, helping operators obtain better quality and more realistic visuals for comprehensive inspections.
When complimented by add-on tools, FiFish can use traditional laser scaling and AI automation to measure underwater dimensions of objects accurately and efficiently, identifying damaged areas in underwater structures, including oil and gas facilities offshore, with high precision, experts at Qysea say.
BUILT FOR PURPOSE
Inspired by the hammerhead shark, E-Go’s biomimetic tail fin stabilises the machine during underwater operations, minimising water flow disturbances. Additionally, the T-shaped fin design enhances operator mobility, enabling swift one-handed handling and rapid deployment, according to Qysea.
The E-Go also features a modular design with detachable components for its motor, lighting, camera, and battery. Module replacement operators can restore the E-Go to its underwater operable state quickly. With extensive expansion capabilities, it can accommodate up to six tools simultaneously.
Qysea claims that the new robotics system will provide a wide range of subsea operations and scenarios with increased power and propulsion.
