Hi there,
I am constructing a ray tracing algorithm , i know that the first step is to develop camera and view plane specifications.
Now is the next step performing rasterization algorithm on image before a BVH tree is constructed so that intersection tests can be performed?
Rather than performing rasterization you might want to calculate bounding boxes for BVH intersections. A suggestion: take a look at the OptiX SDK and OptiX quickstart/programming guide
It is not required, but one can mix these two techniques to achieve interesting effects. You could for example perform ray tracing only for pixels/fragments of a rasterized geometry (less work to do then in case of whole image)…