Software and source code | ||||||||||
Hosting
[AD-IMG] |
Important! Most programs lack documentation. I'll write it shortly, but if there are any questions concerning programs usage (howtos, crashes, bugs etc.) just write me on e-mail.
Gabor filter Gabor filter This program is written to show how the Gabor's filter parameters affect the filter.
Gamma Science The program is written for watching medical images in DICOM format and for
performing various operations for medical analysis, like drawing regions of interests and
accessing their statistics and dynamical changes, watching horizontal and vertical profiles,
measuring distances, using different color scales etc.
Genetic algorithm Genetic algorithm demo program with adjustable algorithm parameters and run statistics.
Graph builder Program for calculation of all shotest paths in directed and undirected graphs and trees. The graph structure can be either loaded using adjacency matrix (the example files are included), or be generated at random. Image categorization This program was written for experiments on using different machine learning
algorithms for automatic images categorization. The training and test images' descriptions
should be created using Image description utility.
Image description This program creates image description using local color and brightness parameters and edge orientation histogram.
There's an ability to set image points sampler to form description from the subset of image pixels.
Also a vector quantization based on the Kohonen's SOM can be used to reduce the resulting description.
The result is written into text and binary files which are created in the program's directory.
Image generator Generates random images filled with random primitives (horizontal and vertical lines, circles).
Kohonen's Self-Organizing Map Program to train Kohonen's Self-Organizing Map using standard and batch algorithms.
Polynomial Bounds Program to compute bounds for complex and real polynomials using classic and modern algorithms.
Separable neuroevolution This program implements neuroevolutionary training of the neural network
to increase dimension of the features space, which should lead to better classification
accuracy according to the Cover's theorem (1965).
Source code: |
Made by Yury Tsoy. 2008-2011. |