Image detection in real time based on fuzzy fractal theory.

Kya Abraham Berthe (1) , Yan Yan (2) , Sounkalo Dembélé (3)

1 ENSup Bamako - École normale supérieure - Bamako
2 USTB - University of Science and Technology Beijing [Beijing]
3 LAB - Laboratoire d'automatique de Besançon

Real time image detection is still a challenge in research. Several methods have been used, but all can be divide in two approaches: the first is based on image field estimation in this case the quality of image is depending on the estimation method. The second is based on electrons collection, the particularity of this approach is that more the collection time is longer, better will be the quality of image. In both of these approaches, the global image should be obtained by assembling the mosaic local image or the visual index of the different point of the image. In this paper we introduce and Hybrid Fractal Fuzzy theory to track image in real time. The error is minimized using RANSAC (Random Sample Consensus) algorithm, by computing the homograph “pixel” union of image. In practice for mobile image a loop can be realize to focus the image in real time, so, we can have and efficient view of the global image in real time, which confer to the propose approach his efficient flexibility.

Mots clés: O^ptimal segmentation; Fractal fuzzy theory; RANSAC algorithm; Image detection; O^ptimal segmentation.

Domaines: Micro et nanotechnologies/Microélectronique

Kya Abraham Berthe, Yan Yan, Sounkalo Dembélé. Image detection in real time based on fuzzy fractal theory.. Conference of the Mali Symposium on Applied Sciences, MSAS'06., Jul 2006, Bamako, Mali. 6 p. ⟨hal-00338577⟩