Rendering and interaction in co-operative 3-D image analysis

Most interactive segmentation methods are defined at a 2-d level requiring sometimes extensive interaction when the object of interest is difficult to be outlined by simple means of segmentation. This may become tedious if the data happens to be a 3-d scene of which a part of the differentiating features of a structure is buried in 3-d contextual information. We propose a concept of extending interactive or semi-automatic segmentation to 3-d where coherence information may be exploited more efficiently. A non-analysed scene may be evaluated using volume rendering techniques with which the user interacts to extract the structures that he or she is interested in. The approach is based on a concept of a co-operative system that communicates with the user for finding the best rendition and analysis. Within the concept we present a new method of estimating rendering parameters if the object of interest is only partially analysed.