Multi-attribute Based Transfer Function Design For Volume Datasets

Volume datasets have been widely used in areas such as medical science,biology,aerography,geology and science simulation.Transfer functions play an essential role in volume data exploration and classification.Transfer functions set optical attributes such as color and opacity to the internal features of volume datasets,and then features in the datasets can be presented to users through volume rendering methods.Thus,a large number of studies in science visualization focus on transfer function design.In this paper,we try to make use of the volume datasets as well as the attributes derived from it to construct transfer functions that are intuitive and user-friendly.Photographic volumes have been increasingly used in medical and biological researches in recent years.The original colors kept in photographic volumes present great opportunities to capture a rich set of information within the dataset for a wide variety of data analysis and visualization applications.Despite years of research,an interactive and user-friendly transfer function is still lacking for photographic volume visualization.The difficulty lies in how to map colors to a space that is convenient and intuitive for users to interactively classify features,i.e.,specifying opacities for voxels.In this paper,we propose a color-based transfer function for intuitive opacity specification of photographic volumes.The color-based transfer function intelligently maps the colors from 3D to 1D,resulting in 256 representative colors that preserve the original colors to the maximun extent.Users can directly classify voxels based on these representative colors similar to the conventional 1D transfer function.Experiments are performed to evaluate the effectiveness of the proposed method,and also demonstrate the intuitiveness and flexibility of the proposed method.Different from scalar volume data,photographic volume datasets are directly captured by means of the modern cryo-imaging systems.The voxels are recorded as RGB vectors,and this makes it difficult to estimate accurate gradients for volume shading and transfer function design,especially when photographic volume datasets are disturbed by noise.In this paper,we propose a robust color gradient estimation method to produce accurate and robust gradients for photographic volumes.Firstly,a high-pass filter is employed to estimate the gradient in a dominant direction and low-pass filters are applied in the orthogonal directions to reduce the efifects of noise.Then,an aggregation operator is applied to estimate both the directions and magnitudes of gradients accurately.Based on the obtained gradients,the shading effects of internal materials are enhanced and the features can be better classified in the transfer function space.At last,the effectiveness of the proposed robust gradient estimation is demonstrated with a large number of experimental rendering results,especially for those noisy photographic volume datasets.Conventional transfer functions are built based on one or more attributes of the data set,such as scalar value and gradient magnitude.Their ability to classify features is limited by the attributes they choose.We propose a hierarchical method for transfer function design,which makes full use of different data attributes.A similarity-based attribute recommendation method is used to choose attributes for transfer functions when users explore the data set or further analysis a specified feature.With our technique,the limitation of one attribute can be made up by others so that users can hierarchically navigate the volume from coarse to fine with different attributes in a convenient manner.