In the previous chapter, I discussed how to take a collection of unstructured volume elements and sample them with a grid of viewing rays. In this chapter, I discuss how to take a collection of samples along a viewing ray and convert the material properties of the volume into the intensity of light that emits from the volume. I do not define a model for the volume in this chapter. I use the model defined in Chapter 1.
When computing the light intensity, I will not consider an infinite viewing ray, but rather I will consider a finite segment of a viewing ray. I assume that the volume properties vary linearly throughout the segment. The adaptive transfer function sampling method introduced in Section 1.3 ensures that the volume properties will vary linearly. Furthermore, I parameterize the volume properties as I do in Chapter 1: with the volume parameters at the front and back of the segment and the length of the segment.
I reviewed several competitive ray integration methods in Section 1.2. However, all of the methods either introduce artifacts (requiring excessive sampling of the volume) or are too computationally intensive to use in real time or interactive environments. The ray integration methods I provide in this chapter have a balance of accuracy and computational complexity.
