Algorithm for extracting points from images: irregular contours

Resumo

This work proposes the development of a software to implement a contour extraction algorithm for
two-dimensional geometries. The algorithm uses different image processing and data analysis techniques to build
an object mask. From the object mask, a finite set of points that describe the image contour is extracted. Next,
the contour is separated by parts with different enumerations delimited by rectangles. All data processed and
extracted from the image can be downloaded by the user in file form. Then, the information from the contour
data is used to generate two-dimensional meshes for finite difference calculations and numerical simulations.
Our automated algorithm generates meshes for irregularly contoured geometries. For this, the mesh elements
in rectangular coordinates are not equally spaced, so that the vertices of the border coincide with the contour of the
geometry obtained by the algorithm. For the development of the algorithm, techniques are used that select a pixel
range of an object to generate the mask. The software uses the Python programming language, together with the
OpenCV library that performs image and data analysis.