Genetic algorithm-based image preprocessing for volume rendering optimization

Volume rendering is a central task in the medical image reconstruction and 3D visualization fields. A new method was proposed in this paper. Instead of optimization of the transfer function or the rendering shader construction, we focus the processing of the original data-set images. In early stage of the volume rendering pipeline, a series of image filters are applied to preprocess the source images. To assist the design of the processing filters, we also adopt the stochastic search algorithm to optimize the filters. Although only the GA method is analyzed, other algorithms can also be used during the process. Experiment results show that this method can optimize the quality of the volume rendered image effectively.