SIVP toolbox for Scilab / News: Recent posts

This version can support Scilab 5.3 on Windows x86 and x64.

Many users reported that SIVP 0.5.1 can not built under Windows. The problem is caused by an error in builder scripts. It has been fixed in SIVP 0.5.2. Sorry for that!

SIVP 0.5.1 can solve some problems with Scilab 5.2. In this release, autoconf and automake scripts are removed. You can just put the files into SCI/contrib/ and then run builder.sce scripts through Scilab to build the toolbox under both Linux and Windows.

SIVP 0.5.0 is improved greatly compared with the previous versions. Some advanced computer vision algorithms, such as face detection, eye detection, foreground detection, object tracking, have been added to this release. Besides, SIVP 0.5.0 can be used in both Scilab 4.X and the newer Scilab 5.x. Enjoy it! :-)

Miss Wu currently is a researcher in the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences. She is working to add face detection, object tracking, background modeling and other functions to SIVP, and her work will improve SIVP greatly.

Good news for Debian users! SIVP has been accepted by Debian. Debian users can install SIVP by a simple command "apt-get install sivp".

In this release, some bugs fixed, and there no great improvement. I also gave a binary release for Windows users. Because there no windows version for 0.4.1 and 0.4.2, Windows users are encouraged to update to this new release.

SIVP won second prize in Scilab Contest of China 2006 (http://www.nlict.zju.edu.cn/scilab). The reward is 1000RMB. ^-^

The release fixed the install path error in version 0.4.0. From now SIVP will be install in path SCI/contrib/sivp-@VERSION@, and a newer version will not overwrite the older oversion.

<a href="http://www.nlict.zju.edu.cn/scilab/index_en.html" target="_blank">Scilab Contest 2006</a> is sponsored for Chinese students and technicians to improve the ability to study and manage Scilab. SIVP 0.4.0 has been submitted to the contest. We hope we can win a prize.(^-^) We would like to see more people know Scilab and use Scilab.

In this new release, many new functions were added including video write functions (but video write functions are still not robust enough). Now the number of functions reaches to 46. Some critical bugs were found and fixed.

The following functions added in this new release: edge, imadd, imsubtract, immultiply, imdivide, imcomplement, imlincomb, imfinfo, imnoise, hsv2rgb, rgb2hsv, rgb2ycbcr, ycbcr2rgb, ntsc2rgb, rgb2ntsc, avifile, addframe.

In my two-week summer vacation, I worked on SIVP. Some critical bugs were found and fixed, and the following functions added:
edge
imadd
imsubtract
immultiply
imdivide
imcomplement
imlincomb
imfinfo
imnoise
hsv2rgb
rgb2hsv
rgb2ycbcr
ycbcr2rgb
ntsc2rgb
rgb2ntsc

You are welcome to try SIVP.

Enjoy it! ^-^

# SIVP 0.3.0 is really a new version which is more compatible with Matlab Image Processing toolbox. Some functions in previous versions but not be listed in Matlab Image Processing toolbox will be discarded. They can still be used in version 0.3.0, but not listed in man pages.

Features for version 0.3.0:

* Supported image:
channel: 1-channel (binary or grayscale image), 3-channel (RGB image)
data type: boolean, uint8, int8, uint16, int16, int32, double
* Image I/O functions: imread, imwrite
* GUI: imshow
* Video I/O, camera read functions: aviinfo, aviopen, camopen, aviclose, avicloseall, avilistopened, avireadframe
* Image type conversion: im2bw, im2uint8, im2int8, im2uint16, im2int16, im2int32, im2double, mat2gray
* Spatial transformation functions: imcrop, imresize
* Image analysis and statistics functions: mean2, corr2, std2/stdev2
* Image arithmetic functions: imabsdiff
* Linear Filtering and Transforms: imfilter, fspecial ... read more

The logos and layout of the web site have been desinged. Hope all of you like it.

We are still active on this toolbox. Sorry for not updating the web site. In the past months:

1. We won the first prize in Scilab 2005 Contest China, and INRIA Scilab Group invited one of the authors, Shiqi Yu, to Paris for one-week visit.
2. We have migrated the CVS repository over to Subversion. Now you can get the latest source code by command:
svn co https://svn.sourceforge.net/svnroot/sivp sivp... read more