期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：2
页码：1334
DOI：10.15680/IJIRCCE.2017.0502006
出版社：S&S Publications
摘要：Inpainting is the art of restoring lost parts of an image and reconstructing them based on the backgroundinformation. This has to be done in an undetectable way. This will be use for to remove the unwanted objects orunwanted data from the original image but this change is not noticeable by the user. Hence this will be done by usingthe two algorithms namely: Exampler Based Algorithm and Fast Marching Method. Digital Image Inpainting tries toimitate this process and perform the Inpainting automatically. Details that are hidden completely by the object to beremoved cannot be recovered by any mathematical method. Such software has several uses. One use is in restoringphotographs. Another use of Image Inpainting is in creating special effects by removing unwanted things from theimage. Unwanted things may range from microphones, ropes, some unwanted person and logos, stamped dates and textetc. in the image. These parts can then be reconstructed using Image Inpainting. Natural images and photographssometimes may contain stains or undesired objects covering significant portions of the images. The process ofremoving objects from images starts with mask out the undesired object, making the area where the object previouslyoccupies a gap. Then the gap will be filled using graphical techniques. Among the graphical techniques that are used tofill the gap after object removal, two most commonly used are: Image Inpainting and texture synthesis. Exemplar basedtechniques, which cheaply and effectively generate new texture by sampling and copying color values from the source.Exemplar based algorithm is used for removing large objects from digital photographs. The Fast Marching Method(FMM) is used to remove all scratches within the image.
