Conservation of identity when using lossy methods works differently with sound than with images. Will I recognize this sound if I put it through image changes? The sound’s intelligibility survived “convert to binary” (2 bit) and doing an outline of the binary forms. but it did not survive a number of other ways. Erosion was the most fascinating though, as it played every sound the test voice I used was capable of all at once but in the rhythm of speaking.

Conservation of identity when … [read full article]

 

A bit of a lot of things – lots of trial and error. But I settled on a process for now: Paint dot NET is my homebase. a) Have color image. b) Convert it to 1-bit using a little thing I found http://blog.roguecode.co.za/update-for-my-agent-1-bit-image-converter-1bitter/ c) Open in Inkscape d) Control-A (Select All) e)Path:Trace Bitmap “Grays” Turn off Smooth Options: Turn off Smooth Corners f) Path: Simplify (or Control-L) about three times g) File: Export Bitmap. To do: 1) Automate b) – maybe imagemagick “convert” can do it? 2) Maybe shrinking the image to like 64×64 will make thresholding work better? I don’t know. 3) Use potrace instead of Inkscape for tracing. I did and can but as I haven’t mastered potrace at all, I have to learn its options (same problem I have with Imagemagick – I don’t know it well) 4) Find another way to “simplify”. I think potrace can but I don’t know for sure.

A bit of a
[read full article]
 

It was the help of an app called Fourier Anime that got me started. It created rubber-band animation from a standard image. I took one of the frames, simplified it further using Inkscape’s “simplify” to get here. App page for Fourier Anime says it uses this process, which I’d like to figure out to do by hand: 1. Edge Detect Differentiate the luminosity value of the monochrome input image, then we can get the edges where the luminosity are changing significantly. 2. Find Contours Extract contour curves from the edge image one by one. 3. Fourier Transform Contours are the sequences of points. Convert this complex sequences into Fourier coefficients by Fourier transform. 4. Low-pass Filter Leave only the coefficients of low-frequency parts. 5. Inverse Fourier Transform Reconstruct the complex sequences of points from low-pass filtered Fourier coefficients. And draw the curves in the display (parametric representation).

It was the help
[read full article]