Supervised image processing scripts are developed to acquire and plunge textual documents, such as scores or handwritten papers, into abstract asemic spaces. Each text is a unique, unquenchable source of signs, which must be thought of as being thoroughly detached from the original figures of writing to merge into new alphabets.
The whole image matrix is morphologically manipulated with the help of routines, which combine neighbouring clusters of pixels according to certain structuring shapes (lines, squares, disks, etc.), depending on specific elements (points, centres, size, etc.) which get adjusted until a new formal equilibrium is attained.
Scripts and other software segments have been developed, tested and run within Linux GNU Octave, whose mathematics-oriented syntax is mostly compatible with Matlab.
Preliminar image analysis has benefited standard ImageJ plugins.
% fragment of a script
x=imread(‘im.jpg’);
E=2*rand(1);
SE5=strl(‘arbitrary’, E);
V=[2,3];
SE=strl(‘per’, 1, V);
SE1=strl(‘li’, 1, 1);
SE3=strl(‘sq’, 1);
NEHOOD=[1 0 0 0 0 0 0 0 1 1 1 1 1 0, 1 0 0 0 0 1 0 1 0 1 0 1 1 1 0 1 1 1 1];
SE4=strl(NHOOD);
OFFST=[10, -18];
SE2=strl(‘pairwise’, OFFSET);
A=imerode(x, SE1);
A1=imerode(A, SE4);
A2=imerode(A1, SE3);
A3=imerode(A2, SE);
B =imerode(A3, SE5);
imwrite(B, ‘imsaved.jpg’);
Stuttgarter Strukturfonien, Morrisville 2019. ISBN 978-0244793913
