Image Super-Resolution Via Sparse Representation

稀疏逼近 人工智能 模式识别(心理学) 计算机科学 图像(数学) K-SVD公司 图像分辨率 代表(政治) 相似性(几何) 分辨率(逻辑) 面子(社会学概念) 计算机视觉 社会科学 社会学 政治 政治学 法学
作者
Shuicheng Yan,John Wright,Thomas S. Huang,Yi Ma
出处
期刊:IEEE transactions on image processing [Institute of Electrical and Electronics Engineers]
卷期号:19 (11): 2861-2873 被引量:5249
标识
DOI:10.1109/tip.2010.2050625
摘要

This paper presents a new approach to single-image super-resolution, based on sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low resolution and high resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low resolution image patch can be applied with the high resolution image patch dictionary to generate a high resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs, reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle super-resolution with noisy inputs in a more unified framework.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
1秒前
徐不想搞科研完成签到,获得积分10
2秒前
小斌发布了新的文献求助10
2秒前
脑洞疼的应助被小张采纳,获得10
2秒前
Akim的应助被马来自农村的马采纳,获得10
2秒前
爆米花的应助被风趣冬瓜采纳,获得10
2秒前
fantasy的应助被dpcrel采纳,获得10
3秒前
Owen的应助被Qianfan采纳,获得10
3秒前
QQ发布了新的文献求助10
3秒前
阳光马里奥完成签到,获得积分10
4秒前
aabbccwy完成签到,获得积分10
4秒前
yeape发布了新的文献求助10
4秒前
小寒0812完成签到,获得积分10
5秒前
1111111111111完成签到,获得积分10
5秒前
5秒前
小二郎的应助被Charlie采纳,获得10
6秒前
6秒前
6秒前
斯文败类的应助被小迪采纳,获得10
6秒前
彬彬完成签到,获得积分20
6秒前
7秒前
LCY发布了新的文献求助10
7秒前
123完成签到,获得积分10
7秒前
8秒前
8秒前
8秒前
实验陈功完成签到,获得积分10
8秒前
桐桐的应助被居居采纳,获得10
8秒前
8秒前
aabbccwy发布了新的文献求助50
9秒前
科研通AI6.4的应助被西瓜宝宝采纳,获得10
9秒前
爆米花的应助被旧日采纳,获得10
9秒前
上官若男的应助被不想起床采纳,获得10
10秒前
河里的星完成签到,获得积分20
10秒前
10秒前
852的应助被大力寒荷采纳,获得10
11秒前
充电宝的应助被橙橙采纳,获得10
11秒前
初景的应助被影子采纳,获得20
11秒前
11秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
Markov Decision Processes and Reinforcement Learning 700
八旗制度の研究 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8168976
求助须知:如何正确求助?哪些是违规求助? 9561992
关于积分的说明 21764222
捐赠科研通 7760642
什么是DOI,文献DOI怎么找? 3430195
关于科研通互助平台的介绍 2554066
邀请新用户注册赠送积分活动 2474834