Fuzzy Mutual Information-Based Intraslice Grouped Ray Casting

Abstract

The traditional ray casting algorithm has the capability to render three-dimensional volume data in the viewable two-dimensional form by sampling the color data along the rays. The speed of the technique relies on the computation incurred by the huge volume of rays. The objective of the paper is to reduce the computations made over the rays by eventually reducing the number of samples being processed throughout the volume data. The proposed algorithm incorporates the grouping strategy based on fuzzy mutual information (FMI) over a group of voxels in the conventional ray casting to achieve the reduction. For the data group, with FMI in a desirable range, a single primary ray is cast into the group as a whole. As data are grouped before casting rays, the proposed algorithm reduces the interpolation calculation and thereby runs with lesser complexity, preserving the image quality.

Research on Improved Ray Casting Algorithm and Its Application in Three‐Dimensional Reconstruction

“…In contrast, the acceleration method based on software is more flexible and convenient, which can be transplanted between different machines quickly and has wider applicability. Mehaboobathunnisa et al proposed a method of grouping rays projected by similar voxels to reduce the computational complexity of rendering algorithm, but the reconstruction result is not smooth enough due to artifacts [10]. Hadwiger et al proposed the SparseLeap method which is a novel space hopping method and has been proved that it can avoid the problem of unnecessary space debris [11].…”

A Novel Ray-Casting Algorithm Using Dynamic Adaptive Sampling

“…The simulation environment is VS2016 and OpenGL. Table 4 shows the comparison of the six algorithms for human head image rendering, and ICVC-GPU, MVRC2, SigIg+FMI, SHSR-UV, DBRay, and Ray-casting DAS in Figure 5 represent the experimental results of Feiniu Yuan's algorithm [26], Mohammandmehdi Bozorgi's algorithm [32], R. Mehaboobathunnisa's algorithm [35], Roba Binyahibs' algorithm [17], Alec G.Moore's algorithm [36], and ray-casting DAS algorithm proposed in this paper, respectively.…”