Title :
Novel neighbor embedding super resolution method for compressed images
Author :
Tao Yu ; Zongliang Gan ; Xiuchang Zhu
Author_Institution :
Key Lab. of Broadband Wireless Commun. & Sensor Network Technol., Nanjing Univ. of Posts & Telecommun., Nanjing, China
Abstract :
In this paper, we focus on the super resolution (SR) reconstruction for compressed images. Compressed images have low quality by suffering from quantization errors and compression artifacts. We propose a novel learning-based SR method that is based on the local linear embedding (LLE). A new feature vector using the discrete cosine transform coefficients of the norm luminance is proposed. It represents the character of the image patch in the transform domain. Compared with several existing feature selections, this new feature vector can preserve edges better and capture more details. It is more suitable for the compressed image. We also optimize the training set. Experimental results show that our method gets better performance both in objective and subject.
Keywords :
data compression; discrete cosine transforms; feature extraction; image coding; image reconstruction; image resolution; learning (artificial intelligence); vectors; SR reconstruction; compression artifact; discrete cosine transform; feature selection; feature vector; image compression; image patch; learning-based SR method; local linear embedding; neighbor embedding super resolution method; norm luminance; quantization error; super resolution reconstruction; training set; Discrete cosine transforms; Image coding; Image reconstruction; Image resolution; Signal resolution; Training; Vectors; feature selection; neighbor embedding; super-resolution;
Conference_Titel :
Image Analysis and Signal Processing (IASP), 2012 International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4673-2547-9
DOI :
10.1109/IASP.2012.6424982
