Subject: Creating 3D model of an investigated object by it`s 2D representations. Recognition of the object.

Teacher: professor Melnikov Boris Feliksovich.

The work is divided into 4 component parts:
1. Extract edges as a straight lines from each representation of the object.
2. Compare the extracted edges and find knots, i.e. the points where a corner is located on different representations.
3. Using formulas of flats and right lines create 3-dimensional model of the object.
4. Recognize the object using it`s 3D model.

I have realized only the first part of the work by now.

The algorithm is:

Color reduction. Decreasing the number of colors in object representations.
Evolving image areas with similar colors, i.e. the nearer pixel is located to the border of colors, the more it dark. As a result of it I get blurred lines covered edges of the object.
Evolving sharp straight lines and arcs from the blurred lines.