影像融合是影像處理領域裡的一個重要議題，它的主要目標是整合多張給定的影像形成一張比原始影像更具視覺品質的融合影像。本文的目標是改進 Li 和 Zeng 在 [14] 針對局部模糊影像所提出的變分影像融合模型。首先，我們設計一種新的選擇準則用來選取給定的多張圖像的一階梯度信息作為圖像特徵，預期該圖像特徵應該接近目標圖像的梯度向量，然後我們以所取得的圖像特徵做為基礎，提出一個變分圖像融合模型用以融合給定的局部模糊影像，並且引用分離 Bregman 迭代法有效地解決相對應的變分問題。最後，數值例子驗證了所提出的新模型的有效性，同時我們還與 Li 和 Zeng 模型 [14] 的結果進行了比較。;Image fusion is an important issue in the field of image processing. The main goal of image fusion is to integrate several source images into a fused image with a better visual quality compared to the source images. The purpose of this thesis is to improve the variational image fusion model for local-blurred images proposed by Li and Zeng in [14]. First, we design a new selection criterion to select the first-order gradient information of source images as the image feature which is expected to be close to the gradient of the target image feature. We then propose a variational image fusion model using this image feature to fuse the given local-blurred images. Furthermore, the split Bregman iteration is employed to efficiently solve the corresponding variational problem. Finally, numerical examples are given to illustrate the effectiveness of the newly proposed model. Comparisons are also made with the results of the Li-Zeng model [14].
