19-09-2017, 09:54 AM
The system looks for the best match between image signatures and model signatures. Determines whether the same type of model is present in the image that has to be segmented or not. The segmentation performed on the model is reported in the image to be segmented by the application of an affine transformation (translation, rotation and homothety).
The proposed system uses our local descriptor developed to achieve robustness against geometric distortion. In addition, we take advantage of the global descriptor by grouping the original image database by the border histogram descriptor since the local descriptor is not guaranteed in time that matches between pairs of images.
The proposed system uses our local descriptor developed to achieve robustness against geometric distortion. In addition, we take advantage of the global descriptor by grouping the original image database by the border histogram descriptor since the local descriptor is not guaranteed in time that matches between pairs of images.