Computer vision and machine learning
Coronis is mainly focused on research, development and technology transfer activities, specialized in computer vision and machine learning.
We collaborate with a number of companies on developing computer vision techniques applied in industry.
Such techniques lead to a highly automated production process, increasing the production speed, reliability and reducing the production costs.
We provide personalized and complete solutions for:
- Production control based on the position and orientation of the product for grasping, manipulation, etc.
- Quality control at various stages in the manufacturing line that takes into account shape, structure, distribution, colour, texture, etc. of the product, ensuring with a high accuracy that quality standards are met.
- Classification of products based on the criteria provided by you for:
- Product selection
- Customized manufacturing
- Packaging, etc.
We are involved in active collaborations with medical centres, aiming to develop novel techniques to better acquire, process and interpret medical data in various modalities (MRI, X-ray, optical, etc.).
- We are developing new tools in the field of medical imaging, combining our experience in computer imaging with that of our various collaborators in optical systems and medical diagnosis, etc.
- Our personnel has experience in research of image processing and analysis techniques applied on Magnetic Resonance Imaging, X-ray/Computed Tomography and Ultrasonography, and fusing such information in order to obtain more reliable medical diagnosis.
- Data-mining and statistical analysis techniques, in which we have extended experience, were successfully applied in improving the early diagnosis process of potentially terminal diseases.
Our extended experience in techniques related to mapping and reconstruction provide the necessary basis for applications such as:
- Mosaicing and 2D mapping of visual, acoustic, etc. data in order to obtain high-detail composite maps.
- 3D mapping of visual data obtained from one or multiple cameras with scientific such as geographical and geological purposes, tourism, gaming, etc.
- 3D reconstruction of objects, at big scale eg. building reconstruction or at small scale, for design, prototyping and quality control, etc.
In the latest years we have focused our R&D interest, among others, towards novel technologies with applications in image understanding and semantic classification.
Our interest was materialized in in both research projects, pushing the envelope of the state-of-the-art; and collaborations with industry, where research ideas were implemented and tested.
The number applications for such techniques grows larger by the day, however, we should underline here:
- Automatic classification, detection, identification of places, characters, and events in multi-media videos for TV production, augmented reality, etc.
- Real-time analysis of surveillance vaideos and detection of characters, vehicles, suspicious behaviour, events, etc.