Unsupervised Image Histogram Peak Detection Based on Gaussian Mixture Model

Image histogram peak detection is a fundamental technique in digital image processing and relative areas. It has been found that Gaussian mixture model is an effective method to obtain the histogram peaks. However, how to set cluster centers and fit truncation data remain problems that deserve to be explored further. To solve the latter problem, this paper proposes a method consisting of data prediction, unsupervised data fitting and peaks acquisition. Extensive experiments are carried out to demonstrate the performance, and the results prove that our method can improve stability, deal with truncation data, and adaptively find histogram peaks.

[1]  R. Redner,et al.  Mixture densities, maximum likelihood, and the EM algorithm , 1984 .

[2]  Nizar Bouguila,et al.  Bayesian learning of finite generalized Gaussian mixture models on images , 2011, Signal Process..

[3]  Nizar Bouguila,et al.  Infinite Generalized Gaussian Mixture Modeling and Applications , 2011, ICIAR.

[4]  Miin-Shen Yang,et al.  A robust EM clustering algorithm for Gaussian mixture models , 2012, Pattern Recognit..

[5]  José Manuel Rebordão,et al.  An amplitude segmentation method based on the distribution function of an image , 1984, Comput. Vis. Graph. Image Process..

[6]  Nizar Bouguila,et al.  Unsupervised keyword spotting using bounded generalized Gaussian mixture model with ICA , 2015, 2015 IEEE Global Conference on Signal and Information Processing (GlobalSIP).

[7]  N. Zhang,et al.  Multiscale roughness measure for color image segmentation , 2012, Inf. Sci..

[8]  D. Brillinger.  The asymptotic distribution of the Whittaker periodogram and a related chi-squared statistic for stationary processes , 1974 .

[9]  S. Sastry,et al.  Liquid crystal textural analysis based on histogram homogeneity and peak detection algorithm , 2012 .

[10]  Hui Zhang,et al.  Bounded generalized Gaussian mixture model , 2014, Pattern Recognit..

[11]  R. W. Wedderburn.  Quasi-likelihood functions, generalized linear models, and the Gauss-Newton method , 1974 .

[12]  Chen Yin.  Automatic low light level image enhancement using Gaussian mixture modeling , 2015 .

[13]  Anil K. Jain,et al.  Unsupervised Learning of Finite Mixture Models , 2002, IEEE Trans. Pattern Anal. Mach. Intell..