Description

In this work, technique is proposed to identify dust particles from the camera image. The dust particles on the image are the small noise on the image which is raised when the picture is clicked and at that time dust is there on the lens of camera. In the previous work, salient based bottom up method is applied which will analyze various pixels of the image. The bottom up method will generate salient object on the basis of similarity of the pixels in the image. The pixels which are similar have one matrix and pixels which have dissimilarity have different matrices. To improve accuracy of dust particle detection, ITTI-KOCH algorithm is used in the proposed work. In the algorithm, saliency is designed in which properties of the image is analyzed and pixels which have dissimilarity are put into different matrices and that matrices pixels are marked as noise in the image.