No reference image blurriness assessment with local binary patterns |
| |
| Affiliation: | 1. School of Electronic Information Engineering, Tianjin University, Tianjin, China;2. Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China;3. Santa Clara University, Santa Clara, CA 95053, USA;1. Department of Electrical, Biomedical and Mechatronic Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran;2. Department of Computer and Information Technology Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran;1. Dip. di Informatica - University of Verona, Strada Le Grazie, 15 - Verona, Italy;2. DPIA - University of Udine, Via delle Scienze, 208 - Udine, Italy;1. Department of Radiological Technology, Tokushima Bunri University, 1314-1 Shido, Sanuki-city, Kagawa, 769-2193, Japan;2. Department of Radiology, Otsu City Hospital, 2-9-9, Motomiya, Otsu-shi, Shiga, 520-0804, Japan;3. Division of Health Sciences, Graduate School of Medical Sciences, Kanazawa University, 5-11-80 Kodatsuno, Kanazawa, 920-0942, Japan;1. Joint Laboratory of High Speed Multi-source Image Coding and Processing, Xidian University, Xian 710071, China;2. Southwest University of Science and Technology, Mianyang 621010, China;3. GuangDong University of Technology, Guangzhou 510006, China |
| |
| Abstract: | In this paper, we put forward an effective and efficient no reference image blurriness assessment metric on the basis of local binary pattern (LBP) features. In this proposal, we reveal that part of the LBP histogram bins present monotonously with the degree of blurriness. The proposed method contains the following steps. Firstly, the LBP maps of an input image are extracted with multiple radiuses. And then, the frequency of pattern histogram is analyzed before part of bins are chosen as the features. In addition, we also take the entropy of these bins as another feature. Finally, we learn the extracted features to predict the image blurriness score. Validation of the proposed method is conducted on the blurred images of LIVE-II, CSIQ, TID2008, TID2013, LIVE3D IQA Phase I and LIVE3D IQA Phase II. Experimental results demonstrate that compared with the state-of-the-art image quality assessment (IQA) methods, the proposed algorithm has notable advantage in correlation with subjective perception and computational complexity. |
| |
| Keywords: | Blurriness/sharpness Image quality assessment (IQA) No reference (NR) Local binary pattern (LBP) |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
