62 lines
3.0 KiB
Markdown
62 lines
3.0 KiB
Markdown
# README
|
|
|
|
### Short explenation of the c# .exe
|
|
|
|
|
|
|
|
1. Choose the algorithm u want to calculate on and press "start" or "enter"
|
|
|
|
2. Choose the number of pixels you want to go for. For more than 20,000 pixels, please use the c .exe, because
|
|
this version is not optimized for high number of pixels. Only for visualising small amount of pixel results.
|
|
|
|
3. Clears the screen and reprints the actual values of your picture or predefined values
|
|
|
|
4. Choose the learnrate and window size you want, learnrate has to be over 0.0 and under or even 1.0.
|
|
The window size has to be an integer.
|
|
|
|
5. Loads a picture, it has to be BMP, GIF, EXIF, JPG, PNG or TIFF
|
|
|
|
6. Check this for generating outputfiles in the same direktory you started your .exe in
|
|
Outputfiles are : weights.txt, weights_after.txt, localMean|directPredecessorer|differentialPredecessorer.txt, results.txt
|
|
|
|
7. Chart where the algorithems output is displayed
|
|
|
|
|
|
|
|
|
|
1. Displays the average error of the calculated algorithm. The first "window size" (M) pixels are ignored for this
|
|
calulation, because the not updated weights are causing wrong calculations. You can see it by entering more than
|
|
20 into the window size.
|
|
|
|
2. Displays the variance of the error of the calculated algorithm. Same as average error, first "window size" (M) pixels are ignored.
|
|
|
|
3. The history of graphs displayed in the chart, it is possible to have graphs of all 3 LMS algorithems in the history.
|
|
There is no way to hide graphs, simply use "Clear" and then calculate again.
|
|
|
|
|
|
### Explanation of the ANSI C/C++ bin
|
|
|
|
Choose an input file `$ CPP_NLMS -i <inputfile>`
|
|
Any __PPM__ file can be used as input.
|
|
|
|
|
|
|
|
The Output contains the default parameters for window size, color channel, seed for randomization, randomly generated weights and outcome of the nlms functions.
|
|
Parameters can be changed. Please use `$ CPP_NLMS -h` so you can see an overview of the given parameters.
|
|
|
|
|
|
|
|
Logs are created by default for:
|
|
+ input data (pixels of __PPM__ file)
|
|
+ created weigths
|
|
+ each nlms function.Predicted value, input value and error value as well as square mean and deviation
|
|
+ updated weights for each nlms calculation
|
|
|
|
`CPP_NLMS -i <inputfile> -g true` prints a graphical overview to `graphResults.html`. The output is an __SVG__ embedded in html.
|
|
|
|
|
|
|
|
|
|
|
|
|