This paper describes an analysis/synt hesis simulation procedure for the generation of realistic infrared images of outdoor scenes. The approach is based on combining synthetic scenes with real images using segmentation algorithms and region analysis. At the analysis stage, the image is segmented into homogeneous regions. Each surface region is then attributed material properties and its orientation is computed. Next, a texture signal is extracted from the original image.At the synthesis stage, thermal models are used to predict surface temperatures and to compute the radionietries of the thermal image. Finally, the extracted texture signal is combined with the synthetic radiometries. Temporal variations of the texture signal are also studied in order to generate the same scene at a different time of day.
This paper presents an algorithm for the segmentation of an image sequence into spatio-temporal regions. This algorithm combines region growing and region tracking. The proposed approach relies on the temporal coherence of the image sequence and no correspondence analysis is needed. The algorithm is applied to image sequences which have been acquired with a moving camera.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.