- Project description
Oceans and seas offer numerous methods of utilization from the coast to the deep sea. They have always been used as routes of transport for worldwide freight traffic and are also a source of high-quality food. Energy and mineral resources in the deep sea are, economically, very attractive.
To sustainably utilize the ocean, technologies are needed that take into account demanding subsea conditions such as high pressures, visibility and currents under water as well as the rough climate above water.
We develop visual computing applications for a broad range of users, from aquaculture to marine mining, based on the special requirements underwater technology needs to fulfill.
Underwater Image Processing
In the submarine world, limiting conditions need to be considered that are different than those ashore: severe interference caused by floating particles and difficult lightning conditions strongly affect the quality of underwater images. With advanced underwater image processing, it is possible to determine the type and size of living organisms to a great extent. Fraunhofer IGD develops sophisticated computer vision algorithms for image enhancement, detection, segmentation and classification of objects and even for the reconstruction of full 3D structures.
Those algorithms solve many tasks in real time and are available as software modules for desktop applications. Alternatively, they can be offered as embedded systems with minimal energy consumption for smart cameras or in underwater vehicles.
SynaptiCam
SynaptiCam is a smart camera for underwater usage. It has been developed for observing or monitoring specific underwater situations, such as the behavior of fish or suspicious objects on the sea floor. Operating on an embedded Linux system, Fraunhofer’s Deep Vision Frameworks enables the detection, classification and segmentation of objects in real-time.
Our SynaptiCam combines low power consumption, small size and easy connection of different camera units with high performance and flexibility for the processing of single images or video streams. Equipped with an optional GSM/LTE module, the result of processing can be provided in real-time.
Technical features:
- Low power consumption
- Full-HD resolution
- On-device image processing with directly controlled algorithms
- Bandwidth-adaptive live view
- Up to 3 camera modules
Potential applications:
- Detection of underwater objects, e.g. fish
- Classification of different fish species
- Segmentation of object forms
Underwater Video Processing
Underwater cameras play an important role in the inspection of underwater buildings, in the detection of unexploded ordnance on the sea floor, and in the monitoring of ROV operations.
Because of the optical characteristics of water, the quality of underwater videos is mostly very poor. Visual inspection by ROV operators, as well as automatic extraction of information using software tools, is hindered by those characteristics.
Fraunhofer IGD offers applied research and ready-to-use technology in the area of underwater video processing. One of the available solutions is a system which is able to increase the quality of such videos in real time on an embedded system using color correction and increased contrast in the images.
Key features of image improvement
- Real-time capability
- Full HD support
- Embedded system
- Low power consumption
- Digital camera support (e.g. ethernet, SDI)
Application examples
- Integration into underwater cameras
- Integration into ROV control panels
- Stand-alone tool for processing of recorded video
Automatic information extraction
Using the quality-improved images, Fraunhofer IGD is also offering additional software modules for high-level image processing and computer vision for:
- Object recognition
- Measuring distances
- 3D reconstruction of structures