Geometrical distortion correction using enhanced feature extraction

With the development in innovation, therapeutic field likewise utilizes distinctive sorts of machine to obtain the pictures for determination. These sensors procure data and make it as pictures. At some point these pictures are influenced by some bending like barrel and pincushion. This is on the grounds that postulations sensors have an imprint concentrate on either focus or edge. As it spotlights on one point so it can't be evacuated however can be remedied subsequent to obtaining tests. Various techniques have been utilized to right this sort of twisting. The past work which we took under thought was to gather data by removing some composition highlights utilizing that element to characterize the picture which will give the right data at some kind of point. Thus, there is still need to enhance the outcomes. Along these lines, in this work Feature extraction process in upgraded by utilizing two or more elements and separating instrument before highlight extraction. At that point gather data by characterizing them utilizing neural classifier. The performance of this proposed scheme is calculated in terms of accuracy which is approximately 94%. It means Distortion of both types is highly corrected.