Societal needs, imaging trends

Over the last 5 years the use of images and video in real-life applications has exponentially increased. There has also been a shift from simple recording and playback towards highly sophisticated image processing. This shift has happened in the complete image processing chain, ranging from acquisition, over quality improving processing and reproduction. Enabled by advances in computer vision, products involving highly-sophisticated image analysis technology have appeared in our daily lives.

Project Rationale

In many application domains all over the world there is a trend from "single-view imaging" towards "multi-view imaging". Some examples are the introduction of 3D/4D imaging in Healthcare, High Definition TV and beyond in Broadcasting, multiple connected Surveillance cameras for monitoring a scene. These applications have in common that they all involve an explosive growth in the generation of images and the need for real time handling of images and data. The question arises how to deal with this growth of complexity. The technological concepts and hardware architectures of existing systems were developed in the past when the amount of data and systems were significantly less complex.

Breakthrough technologies are needed to make imaging applications, such as the ones described above, handle the more and more demanding requirements concerning image quality and reliability and speed of image analysis:

  • Image acquisition needs to become much more intelligent about what data to acquire (e.g., automatic region of interest detection) and how to optimally image it based on the image content. In view of the high data rates, this content-aware imaging has to be built into the acquisition devices.
  • Multi-view video processing requires sophisticated inter-camera calibration tools and sophisticated algorithms to merge many views into a more useful, non-redundant higher-level representation, such as a panorama, or a 3D representation.
  • Acquisition devices will also need to become more context-aware: devices should be aware of the existence of other cameras and contribute to the joint optimization of the imaging system in which they are used based on external feedback.

PANORAMA will deliver solutions for applications in medical imaging, broadcasting systems and security & surveillance, all of which face similar challenging issues in the real time handling and processing of large volumes of image data:

  • A huge increase in amount of data that needs to be processed
  • An increase in the complexity of image processing and image analysis algorithms
  • Tight real time requirements for image based control systems
  • Representation of "holistic" image data.

PANORAMA will provide autonomous image acquisition, tightly coupled to the image sensor by research, development and demonstrating generic breakthrough technologies and hardware architectures for a broad range of imaging applications. Solutions will be delivered that require the real time handling and processing of large amounts of image data and develop new CMOS image sensor silicon supporting the autonomous image acquisition. In all these application domains the solutions will support the user by introducing more intelligent components to the imaging system, such that the user can work more efficient and can give more attention to the primary task of the application.

PANORAMA is a research project of the ENIAC Joint Undertaking (JU) and is co-funded by grants from Belgium, Italy, France, the Netherlands, and the United Kingdom.