The key idea: Experimentation!

 

 


 

SIPER is an educational tool for demonstrating signal processing techniques, integrated with MATLAB and with embedded evaluation boards. It can also be used as an analysis tool for research purposes.

 

 

·        Allowing experimentation with key parameters.

 

·        Both intermediate and final results are shown.

 

 

 

 

                                                

 

 

A variety of image processing techniques

 

 

 


 

 

 

 

 

Color representation

RGB, YCbCr, HSV, LAB, …

 

 

Morphological operators

erosion, dilation, opening, closing, top-hat, black-hat

 

 

Analysis

histogram, edge detection, thresholding, segmentation

 

 

Enhancement

filtering, histogram equalization, white balance, crosstalk cancellation

 

 

Resizing

in spatial domain, in DCT domain, seam carving

 

 

Compression

intra prediction, JPEG

 

 

Texture synthesis

chaos mosaic, quilting, graph-cut

 

 

Tools

noise addition, quality evaluation (PSNR, MSE, SSIM)

 

 

 

 

 

 

 

A modular plug-in architecture

 

 

 

·        Allowing easy addition of new techniques.

 

·        A plug-in abstraction layer supplies GUI and I/O services to the plug-in so only a minimal programming knowledge is required in order to write a new plug-in.

 

·       New plug-ins can be written in C, C++, MATLAB® or any ARM-based embedded system.

 

 

 

 

 

 

 

A full demonstration system

 

 


 

 

The figure to the right depicts an example layout of a SIPER demonstration system. A host PC is running SIPER that interconnects to C and MATLAB® host plug-ins and to an ARM+DSP Beagleboard development board from Texas Instruments running embedded ARM and DSP plug-ins.

 

 

 

 

 

 

 

 

 

  

Download SIPER 2.0.0

 

 

 

 

 

 

For any support issue regarding SIPER, please send e-mail to: siper@sipl.technion.ac.il .

 

 

SIPER is developed by the Signal and Image Processing Laboratory (SIPL), Technion – Israel Institute of Technology.

SIPER is intended for non-commercial use and is licensed under the Creative Common 3.0.