| |
Digital signal processing is one of the fastest growing field and
has wide range of application areas ranging from speech processing,
wireless, image processing, geophysics, oil exploration, detection
of nuclear explosions, medical instrumentation and others. Everyones
life on the planet is touched by dsp in one form or other. When
people use phones it has a dsp inside, which does significant amount
of signal processing, though the gadget looks overly simple. Similar
experience users have when they watch LCD TV with crystal clear images
and pure sound. This again has been possible by various algorithmic
improvements that has happened over past few years in the area's of
compression, transmission, storage and display.
Our industry experience of 12 years is mostly in the area of digital signal processing with special emphasis on real time image/video
processing as an application area. Since we come from VLSI background we have more experience on those algorithms which can be implemented on silicon.
In digital signal processing we specialize in all aspects of signal processing, to mention a few:
1. Various kinds of filter design by windowing approach and
remez exchange algorithms, coefficient quantization,
structure selection, finite precision modelling and
analysis (SNR analysis), RTL implemetation and verification.
We can design and implement different kinds of filters like
FIR, IIR, half band, sinc filters etc.
2. Multirate processing: Rate change by means of decimation and
interpolation. Polyphase implementations. Low power designs.
3. Digital decimation and interpolation filters for sigma-delta
codecs. Sigma-delta modulator modelling.
4. System identification.
5. And many more........inclusive of DFT and FFT.
Since DSP is a powerful technique to solve problems in many area's, our
area of expertise is real time image and video processing. But these
techniques are very common and can easily be applied to other domains
like digital communication and audio processing.
Our expertise in digital image/video processing includes:
1. Various kinds of noise reduction
- Impulse noise reduction
- Gaussian noise reduction (3D motion adaptive)
- Mosquito noise reduction
- Deblocking
- Gaussian noise estimation
- Gaussian noise generation
- wavelet domain de-noising.
2. 3D-motion adaptive video/film de-interlacing.
3. Image scaling which includes upconversion and downconversion.
(Poly-phase scaling approach).
4. Color quantization (also known as dithering or error diffusion).
5. Color space and data format (444, 422, 411, 420) conversions.
6. Image Sharpening.
7. Digital video encoding and decoding (NTSC and PAL).
8. Other image processing algorithms like motion detection, edge detection scene cut detection etc.
|
Image Processing
|
| |
