Purpose: This paper presents a new embeddable method for polyp detec- tions in both endoscopic video images and wireless capsule endoscopic images - WCE. Methods: This approach consists rst in extracting possible polyps within the image using some geometric considerations about the shape of the polyps, and then in re n- ing the obtained regions of interest by a boosting-based method using texture features. The performances of the method have been evaluated on a large dataset of 300 polyps, and 1200 non-polyps images. Results: The results show interesting performances of the boosting-based classi cation with a sensitivity of 91%, a speci city of 95% and a false detection rate of 4.8%. These performances are similar to those obtained by Bernal et al. on the same dataset, but the proposed approach reduces the complexity of the hardware implementation for embedded use in WCE or videoendoscope.