透過您的圖書館登入
IP:18.218.70.93
  • 學位論文

藉由學習稀疏性表達特徵的方式偵測圖中視覺顯著區域

Learning Sparse Feature Dictionary for Saliency Detection

指導教授 : 陳煥宗

摘要


圖中顯著性的偵測在計算機視覺研究領域中變得越來越流行。在這篇論文中,我們提出了一種新的方法來產生顯著性偵測的逼近結果圖。其基本思路是使用稀疏編碼係數當作是一種特徵來組成我們要的結果。我們的方法包括兩部分:訓練步驟和測試步驟。在訓練步驟,我們利用在圖片中找到的特徵以及利用人眼注視這張圖的資料來產生圖片相關的字典和如何從稀疏係數轉換成結果;在測試步驟中,給一張新的圖片,我們可以藉由基於特徵的字典得到其相應的稀疏編碼,然後生成結果。我們使用洗牌式AUC以及兩個圖片資料庫來評估我們的研究結果並證明我們的方法可以利用稀疏係數來學習和產生顯著性偵測的結果圖。

並列摘要


Saliency detection becomes more and more popular in computer vision research field. In this thesis we present a new method to generate the saliency map. The basic idea is to use the sparse coding coefficients as features and find a way to reconstruct the sparse features into a saliency map. Our method consists of two parts: training step and testing step. In the training step, we use the features generated from images and the fixation values from ground-truth fixation map to train the feature-based dictionary for the sparse coding and the fixation-based dictionary for converting the sparse coding to a saliency map. In the test step, given a new image, we can get its corresponding sparse coding from the feature-based dictionary and then generate the result. We evaluate our results on two datasets with the shued AUC score and demonstrate that our method gives an efficient sparse coding learning and combination for saliency detection.

並列關鍵字

Saliency Sparse Coding Dictionary Feature fixation Learning

參考文獻


[1] Codruta Orniana Ancuti, Cosmin Ancuti, and Philippe Bekaert. Enhancing by
saliency-guided decolorization. In CVPR, pages 257{264, 2011.
[2] Shai Avidan and Ariel Shamir. Seam carving for content-aware image resizing.
ACM Trans. Graph., 26(3):10, 2007.
[3] Ali Borji and Laurent Itti. Exploiting local and global patch rarities for saliency

延伸閱讀