Abstract
Vector quantization (VQ) has been successfully used in data compression and feature extraction areas. Codebook design is the essential step of VQ. The K-means algorithm is a famous data clustering technique which is also an efficient codebook design scheme. The main disadvantages of K-means algorithm lie in that the initial cluster centroids greatly affect the convergence speed and the final clustering performance. In the past two decades, many codebook initialization techniques have been proposed. However, most of these techniques do not make full use of the features of the training vectors, and some techniques require high extra computational load. This paper presents an efficient and simple technique for the conventional K-means algorithm based on feature classification and grouping. Firstly, all training vectors are classified into sixteen categories based on a two-level classifier including an edge classifier and a contrast classifier. Then, the training vectors in each category are sorted based on their norm values and divided into groups. Each group has the same size, and the centroid vector of each group is calculated as an initial codeword. Experimental results show that, compared with several typical initialization techniques, our technique can obtain a better codebook along with a faster convergence speed in a shorter time.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Alkhalaf S, Alfarraj O, Hemeida AM (2015) Fuzzy-VQ image compression based hybrid PSOGSA optimization algorithm. IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), pp 1–6
Chang CC, Li YC, Yeh JB (2006) Fast codebook search algorithms based on tree-structured vector quantization. Pattern Recogn Lett 27(10):1077–1086
Chen SX, Li FW (2010) Initial codebook method of vector quantisation in Hadamard domain. Electron Lett 46(9):630–631
Chen SX, Li FW, Zhu WL, Zhang TQ (2008) Initial codebook algorithm of vector quantization. IEICE Trans Inf Syst E91-D(8):2189–2191
Equitz WH (1989) A new vector quantization clustering algorithm. IEEE Trans Acoust Speech Signal Process 37(10):1568–1575
Filho RME, Gomes JGRC, Petraglia A (2016) Codebook calibration method for vector quantizers implemented at the focal plane of CMOS imagers. IEEE Trans Circuits Syst Video Technol 26(4):750–761
Gersho A, Gray RM (2012) Vector quantization and signal compression. New York, Springer Science & Business Media
Hong YWP, Huang LM, Li HT (2017) Vector quantization and clustered key mapping for channel-based secret key generation. IEEE Trans Inf Forensics Secur 12(5):1170–1181
Horng MH, Jiang TW (2011) The artificial bee colony algorithm for vector quantization in image compression. The 4th IEEE International Conference on Broadband Network and Multimedia Technology (IC-BNMT), pp 319–323
Kang GY, Guo SZ, Wang DC, Ma LH, Lu ZM (2013) Image retrieval based on structured local binary kirsch pattern. IEICE Trans Inf Syst 96(5):1230–1232
Katsavounidis I, Kuo CCJ, Zhang Z (1994) A new initialization technique for generalized Lloyd iteration. IEEE Signal Process Lett 1(10):144–146
Lai JZC, Liaw YC, Liu J (2008) A fast VQ codebook generation algorithm using codeword displacement. Pattern Recogn 41(1):315–319
Lee D, Baek S, Sung K (1997) Modified K-means algorithm for vector quantizer design. IEEE Signal Processing Letters 4(1):2–4
Leitao HAS, Lopes WTA, Madeiro F (2015) PSO algorithm applied to codebook design for channel-optimized vector quantization. IEEE Lat Am Trans 13(4):961–967
Linde Y, Buzo A, Gray RM (1980) An algorithm for vector quantizer design. IEEE Trans Commun 28(1):84–95
Mirzaei B, Hossein NP, Dariush AM (2014) An effective codebook initialization technique for LBG algorithm using subtractive clustering. Iranian Conference on Intelligent Systems (ICIS), pp 1–5
Oliveira FDVR, Haas HL, Gomes JGRC, Petraglia A (2013) CMOS imager with focal-plane analog image compression combining DPCM and VQ. IEEE Trans Circuits Syst I Reg Papers 60(5):1331–1344
Pal AK, Sar A (2011) An efficient codebook initialization approach for LBG algorithm. Int J Comput Sci Eng Appl 1(4):72–80
Paliwal KK, Ramasubramanian V (2000) Comments on modified K-means algorithm for vector quantizer design. IEEE Trans Image Process 9(11):1964–1967
Qin C, Hu YC (2016) Reversible data hiding in VQ index table with lossless coding and adaptive switching mechanism. Signal Process 129(1):48–55
Somasundaram K, Vimala S (2010) A novel codebook initialisation technique for generalized Lloyd algorithm using cluster density. Int J Comput Sci Eng 2(5):1807–1809
Vasuki A, Vanathi P (2006) A review of vector quantization technique. IEEE Potentials 25(4):39–47
Xiong H, Swamy MHS, Ahmad MO (2004) Competitive splitting for codebook initialization. IEEE Signal Processing Letters 11(5):474–477
Yang PY, Tsai JT, Chou JH (2016) PCA-based fast search method using PCA-LBG-based VQ codebook for codebook search. IEEE Access 4:1332–1344
Yu FX, Luo H, Lu ZM (2011) Colour image retrieval using pattern co-occurrence matrices based on BTC and VQ. Electron Lett 47(2):100–101
Zhao WW, Lai LF (2016) Distributed detection with vector quantizer. IEEE Trans Signal Inf Process Netw 2(2):105–119
Acknowledgements
This work was supported partially by the financial support from the National Nature Science Foundation of China under grants No. 61633019 and No. 61272020 and Zhejiang Provincial Natural Science Foundation of China under grant No. LZ15F030004 and Ningbo Science &Technology Plan Project (2014B82015).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Wang, L., Lu, ZM., Ma, LH. et al. VQ codebook design using modified K-means algorithm with feature classification and grouping based initialization. Multimed Tools Appl 77, 8495–8510 (2018). https://doi.org/10.1007/s11042-017-4747-1
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-017-4747-1