ANGLE ORIENTED BASED IMAGE ANALYSIS USING L-AXIAL SEMI-CIRCULAR MODEL

In every picture recognition system, the input image is selected and compared with the database image for the matching process. If Input image size is not equivalent to database size, the input image is to be resized to match with the size of database image. Before going to compare the pose of the image in both input and database images by using Normalization and Feature Extraction. Using angle orientation, which is alternative of normalization approach for identification system, is considered. In this regard, for a rotated version of another image of face, where the face is in the dataset, though the image that was rotated is not in the database. Our proposed model, angle oriented pose recognition system (multiple images) classifies the image dataset obtained from the input devices using L-axial based semicircular model. To check whether the L-axial based angle oriented input image dataset belongs to same person in the database image or not by using the Test for Randomness and Goodness-of-fit with the help of Chi-Square and Rayleigh test. During the Identification Process two cases arise. In the first case, matching process, the angle oriented images comparing with the images of the database. In the second case, let another image of the face comparing with rotated image which is not in the dataset. To handle the drawbacks which are mentioned in the experiment, we used angle orientation technique, which is addressed as two cases 1) various angles of input images i.e. 0º-90º namely angle oriented rotations with different scale of the poses. 2) To improve the reliability of performance recognition rate to high and compare with other scale (angle) changes, we solve the orientation problem.