PREDICTION OF IMAGE PARTITIONS USING FOURIER DESCRIPTORS - APPLICATION TO SEGMENTATION-BASED CODING SCHEMES

This paper presents a prediction technique for partition sequences, It uses a region-by-region approach that consists of four steps: region parameterization, region prediction, region ordering, and partition cr eation, The time evolution of each region is divided into two types: r egular motion and shape deformation, Both types of evolution are param eterized by means of the Fourier descriptors and they are separately p redicted in the Fourier domain, The final predicted partition is built from the ordered combination of the predicted regions, using morpholo gical tools. With this prediction technique, two different application s are addressed in the context of segmentation-based coding approaches , Noncausal partition prediction is applied to partition interpolation , and examples using complete partitions are presented, In turn, causa l partition prediction is applied to partition extrapolation for codin g purposes, and examples using complete partitions as well as sequence s of binary images-shape information in video object planes (VOP's)-ar e presented.