I am going through some ideas that slice a 3D model and fill the area of the resulting polygon contour . Currently they are using the following concepts as follows:
Every resulting line of the polygon after the slicing operation has a parent triangle, as the original model is made of triangle .
Each of the triangle has a face normal.
During scan conversion, when a scanline at a particular height intersect any of the polygon contour line, the normal information of the parent triangle is utilized to find out if the intersection is happening while getting into the contour or getting out of the contour. All these information are used to generate the region that are to be filled.
Then there is another way called the even-odd filling rule. I do not think it is necessary for me to describe it here. I believe that all of you are very well aware of it.
My question is which one of it is better option to go with in terms of correctness.