Computer-assisted cel animation: post-processing after inbetweening

Although 2D animation production today has become more efficient with the help of improved computer technology, it is still very labor intensive. The less creative inking/painting and inbetweening processes take a significant portion of the total work. Our Computer-assisted Cel Animation (CACAni) research aims at automating the above two processes to assist animators in producing their work faster and with better quality. In this paper, a post-processing algorithm for generating inbetween frames is proposed. Algorithms for generating inbetweenings, particularly, the Feature-based Matching (MFBA) of keyframes and linear inbetween frame generation are also introduced. Images constructed with MFBA may have gaps (artefacts) due to the lack of smoothness among adjacent displacement vectors as well as the discrete feature of rasterization. With the guidance of displacement vectors, our algorithm classifies the gaps into several categories and handles them accordingly. The results are compared with those processed by filtering method to show the superiority of the new algorithm.