低解析度的影像放大後，因為影像缺乏高頻的資訊，可能造成影像在視覺上顯得較為模糊；並且會出現鋸齒狀邊緣的現象。可以藉由低頻的資訊來預測及填補高頻資訊來解決上述的問題。在本論文中，我們提出了一個利用向量量化器作為預測高頻資訊的方法。首先運用LBG演算法及分類向量量化演算法訓練出低頻對應高頻的碼書。將所輸入的低頻影像找出在碼書中最符合的三個向量碼，乘上三個選定的權重值後使得與輸入影像的誤差值最小，然後利用找出的這幾個低頻向量碼對應到的高頻向量碼建立出預測的高頻資訊。最後，建立後的高頻資訊與放大後的影像相加，就能填補其缺乏的高頻資訊，改善影像模糊的問題。同時，我們提出改良式演算法，進一步降低所提演算法的運算複雜度。實驗結果呈現，與其他方法比較之PSNR數值以及重建後的高解析影像。

The low resolution image may be blurred and jagged edge after enlarged because the lack of high-frequency information. In order to solve this problem, we can predict the high-frequency information to compensate the target enlarged image by making use of the low resolution information. In this thesis, we present a novel algorithm to predict high-frequency information by using Vector Quantization (VQ). Firstly, using LBG algorithm and the classified VQ strategy establishes a codebook for low and high resolution patch. Secondly, using the input low resolution patch to find the best match three code vectors of the established codebook. The selection of the matched three code vectors multiplied by the given weight makes the smallest error compared with an input patch. Furthermore, we use the best matched low resolution patches to reconstruct the high-frequency image with the corresponding high resolution patches. Finally, the reconstructed high frequency image and the input image combine to compensate the loss of high frequency information. Therefore, blur and jag on the enlarged image is reduced. Meanwhile, we also proposed an improved algorithm to reduce the computational complexity in advance. The experimental results show that the performance of PSNR and the reconstructed high resolution image compared with other methods.

[1] Baker, S., Kanade, T.: Hallucinating faces. In: Proc. IEEE Conf. Autom. Face Gest. Recogn., pp. 83-88(2000)
[2] Hong Chang, Dit-Yan Yeung, Yimin Xiong: Super-Resolution Through Neighbor Embedding. Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition(CVPR’04) 1063-69(2004)
[3] Tak-Ming Chan, Junping Zhang, Jian Pu, Hua Huang: Neighbor embedding based super-resolution algorithm through edge detection and feature selection. Pattern Recognition Letters 30 494-502(2009)
[4] Dai, S., Han, M.: Soft edge smoothness prior for alpha channel super resolution. In: Proc. IEEE Conf. Comput. Vision Pattern Recognit (CVPR), pp.1-8(2007)
[5] Michael Elad, Dmitry Datsenko: Example-based Regularization deployed to super-resolution reconstruction of a single image. The Computer Journal vol.52 no.1 (2009)
[6] Farsiu, s., Robinson, M.D., Elad, M., Milanfar, P.: Fast and robust multiframe super-resolution. IEEE Trans. Image Process. 13(10), 1327-1344(2004)
[7] Freeman, W.T., Pazstor, E.C.: Learning low-level vision. Int. J. Comput. Vision 40(1), 25-47(2000)
[8] Hou, H.S., Andrews, H.C.: Cubic spline for image interpolation and digital filtering. IEEE Trans. Acoust. Speech Signal Process. 26(6), 508-517(1978)
[9] Hardie, R.C., Barnard, K.J., Armstrong, E.A.: Joint MAP registration and high-resolution image estimation using a sequence of undersampled images. IEEE Trans. Image Process. 6(12), 1621-1633(1997)
[10] Sung-Ho Hong, Rae-Hong Park, Seungjoon Yang, Jun-Yong Kim: Image interpolation using interpolative classified vector quantization. Image and Vision Computing 26 228-239(2008)
[11] Xiaoguang Li, Kin Man Lam, Guoping Qiu, Lansun Shen, Suyu Wang: Example-based image super-resolution with class-specific predictors. J. Vis. Commun. Image R. 20 312-322(2009)
[12] Jinjun Wang, Yihong Gong: Fast image super-resolution using connected component enhancement. ICME 157-160(2008)
[13] Jinjun Wang, Shenghuo Zhu, Yihong Gong. Resolution enhancement based on learning the sparse association of image patches. Pattern Recognition Letters 31 1-10(2010)
[14] Sun, J., Xu, Z., Shum, H.: Image super-resolution using gradient profile prior. In: proc. IEEE Conf. Comput. Vision Pattern Recognit.(CVPR), pp. 1-8(2008)
[15] Jianchao Yang, John Wright, Thomas S. Huang, Yi Ma: Image Super-Resolution Via Sparse Representation. IEEE transactions on image processing vol.119 no.11 (2010)
[16] Liangpei Zhang, Hongyan Zhang, Huanfeng Shen , Pingxiang Li: A super-resolution reconstruction algorithm for surveillance image. Signal Processing 90 848-859(2010)