Computers and Computer Vision
Recent advances in both mathematical and computational
methods in image processing and analysis, along with the dramatic development
of hardware capabilities, have made it possible to build advanced imaging
and photographic system that can be applied to tasks, considered imaginary
only recently. Consequently, artificial tools for vision are being developed
in a diversity of fields. Many everyday advanced tools are already equipped
with computer vision abilities. The Vision and Images Sciences Laboratory
combines research in human vision with the development of unique methods
for image processing and analysis and techniques for computer vision,
medical imaging, machine learning and more. As an example, an advanced
adaptive gain control (AGC) camera was developed in the lab that resembles
the functioning of the eye and thus gains a wider dynamic range of sensitivity.
Smart cameras that incorporate extensive computational power, implement
real-time vision algorithms.
Fields of interest
The laboratory deals with a variety of aspects
of Vision Systems, which must cope simultaneously with capturing, processing
and comparing images, and taking decisions according to the visual information.
Therefore topics for research and development in the laboratory are:
- Image Enhancement based on understanding the
physical formation process of images, e.g. Imaging through scattering
media, multispectral Imaging, high dynamic range imaging, scene analysis
using polarization cues.
- Development of new methods for pattern recognition
and classification, e.g. learning systems and new optimization methods.
- Enhancement of methods for medical imaging.
- Methods for blind separation of different
types of tissue by means of tissue finger-prints.
- Algorithms for stereo vision.
- Real-time processing of image sequences /
- Fusion of information from different imaging
Examples for projects in the laboratory
- Smart video camera that mimics the human eye's
ability to compensate for uneven illumination in high-contrast scenes
(increasing dynamic range).
- Underwater photography based on specifically
designed image processing methods.
- Object recognition from satellite images based
on state of the art mathematical methods.
- Recognition of brain activity from medical
images, e.g. MRI.
- An autonomic vehicle that navigates by means
of visual information and avoids obstacles.
- Camera based driver assistance systems for
identifying dangerous situations.
- Automatic recognition of offside
in soccer game.
- Automatic filtering of advertisements in TV
- Smart entrance control using face recognition
based on a miniature camera and a PC.
- Vision through clouds: development of methods
for seeing through scattering media.
- Algorithms for automatic face recognition.
Research and Teaching
Pure and applied research in the fields of image
sciences is growing rapidly in the academy in recent years. Fruits of
the research performed both at the Technion have contributed much to the
growth of hi-tech industries. As an outcome, research and teaching activities
in the lab have doubled in a short period of time.
- Medical Centers: TEL HASHOMER, BNEI
ZION, PORIA, RAMBAM, ICHILOV, ...
- The Biomedical Department, Technion
- Department of Medicine, Technion (brain research)
- Industry: i-Sight, GE Medical, ELBIT, ELOP,
- Ministry of Defense
- Ministry of Transportation
- Ministry of Science