AUTOMATIC IMAGE RETARGETING USING SALIENCY BASED MESH PARAMETERIZATION 1 S.Sai Kumar, 2 A.Sudhir Babu Department of Computer Science Engineering PVP Siddhartha Institute of Technology, Vijayawada, Andhra Pradesh, India. 1 E-mail:saikumar_vij@yahoo.com 2 E-mail:asbabu@hotmail.com Abstract— Automatic image retargeting is used for large image.That are to be fit in small size display devices. Without any loss of information.our proposed methods one is saliency based mesh parameterization method is used to retarget the image. Stretch-Based Mesh Parameterization is generating on saliency image. Graph- Based visual saliency is used to easily find out the objects in the image. Mesh Generation on that saliency object part it should be retarget the image. Another method is comparison of seam-carving and Automatic retargeting method. In this we show the same image as differentiate by axes to separate one is loss of information and structure of the image is changed is called seam-carving. Another one is without loss of information and should not change the structure of the image is called retargeted method. Key words:Automatic image retargeting, Mesh parameteriztion,seam-carving. I. INTRODUCTION Now a days mobile devices like cellular phones,PDA’s are increasing more and more.these are all portable display image functionalities. Mobile devices having small screens[1] with different aspect ratio and much more highest resolutions.Those are to be fit in target display. The problem of adapting the images to fit in various trget screens by using image retargeting. A common functioality of image retrgeting is to rescale the original image according to the target screen size. Scaling method has the problem. That is some important portion of the image should be decreased. So visual distortion is introduced. The next method is cropping to retarget the image. It is used to crop the important area of the image. The main disadvantages of cropping are contextual content of the particular image to be re moved. Another method is seam- carving [4] [6]. In this image resizes by iteratively remove the low resolution pixel values. The main drawback of seam- carving is to break the objects of the image. So image structure is changed. The last method for image resize by using mesh parameterization [8]. In this mesh is generation on color images. Where we select the objects from the image is manually by using blue edge. Reduce the edges of the objects and to fit in target size. It also has the disadvantage that is mesh generation on color image is more time taking processes. To find out the threshold values of the image it takes lot of time. Another one is by selecting the objects of the image is manually. To overcome these problems our method is introduced that is Automatic image retargeting [3] using saliency [2] based mesh parameterization. In this we define two areas one is saliency based mesh parameterization and another one is comparison of seam-carving and retargeting methods. Now our proposed method  Initially taken the input color image.  Input image is discriminated by row-column wise.  Calculate the edges by using CANNY EDGE DETECTION algorithm.  Apply saliency to change the color image to Gray- Scale image. Where the objects are found it takes high resolution and it indicates white color, low resolution indicates black color.  The saliency image is to be discriminated by row- column wise. Where the high resolution is found it indicates black discrimination points, low resolution indicates white discrimination points.  Apply Stretch-Based Mesh Parameterization and using Delaunay Triangulation to retarget the image. With out any loss of information.  To produce the retarget image.  Comparison of seam-carving and automatic retargeting methods. II.RELATED WORK The problem of retargeting images to small screens has been considered by many researchers. Almost all image browsers on small-screen devices provide scaling functionality. A large S.Sai Kumar et al. / (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 1 (4) , 2010, 273-279 273