Research On The Technique Of The Seamless Image Mosaic

The purpose of the seamless image mosaics is to make the mosaic image smooth and seamless near the seam and harmonious in the globe intensity through modifying the pixel value; and on the other hand, it is to keep drawing information from the to-be-mosaic images at the most extent. Image mosaics, as an important image processing technique, is widely utilized at present in many application fields such as computer vision,medical image processing,virtual reality and so on. Starting from the pointview of the existing format of the seam in the mosaics image, the main research content in this paper is described as follows.Firstly, we pay a more attention to remove the seams among the to-be-mosaic images including the transition areas. In order to give attention to the two image mosaic evaluation indices of the smooth and distortion simultaneously, an optimal image mosaic method based on fuzzy integral is proposed in this paper. Aiming at the problems resulting from a single-factor evaluation index, the integrative evaluation method of the mosaic image is put forward based on synthesizing the two independent single-factor indices by fuzzy integral which is largely subject to the human feeling. According to the optimal image mosaic method, an optimal weighted function is obtained. With the help of the optimal weighted function, wavelet coefficients are fused at different resolution levels. As a result, a smooth and seamless mosaic image is reconstructed.Secondly, we pay a more attention to remove the seams among the to-be-mosaic images hardly including the transition areas. In order to give attention to both the smoothness of the boundary area near the seam and the coincidence of globe intensity in a resultant mosaic image simultaneously, an image mosaic method based on oriented vector field is proposed in this paper. With respect to the problem that it is very difficulty for the existing image mosaic algorithms to be adaptive for the big gray difference between the to-be-mosaic images, an oriented vector field with the distributional characteristics of to-be-mosaic object is set up at first. And then the resultant mosaic object is updated by interpolating and iterating under the guidance of the vector field. Finally, when the minimum gradient value near the seam is achieved, an ideal mosaic image with the effect of harmony in globe, and local smooth transition can be obtained.Thirdly, aiming at the study of removing the seams in a mosaics image , we summarize the methods of removing different seams in this paper and present the selective rules of the weighted functions when the pixel value are modified adopting the method of the weighted fusion;at the same time,based on objectivity and visual perception we present some evaluated indices and the methods of synthesizing these evaluated indices to obtain the mosaic optimal algorithm according to the feedback.