Robust Gray-Level Histogram Gaussian Characterisation
One of the most utilised criteria for segmenting an image is the gray level values of the pixels in it. The information for identifying similar gray values is usually extracted from the image histogram. We have analysed the problems that may arise when the histogram is automatically characterised in terms of multiple Gaussian distributions and solutions have been proposed for special situations that we have named degenerated modes. The convergence of the method is based in the expectation maximisation algorithm and its performance has been tested on images from different application fields like medical imaging, robotic vision and quality control.
KeywordsImage Segmentation Grey Level Expectation Maximisation Algorithm Image Histogram Medical Image Analysis
- 5.Punam K. Saha and Jayaram K. Udupa. Optimum image thresholding via class uncertainty and region homogeneity. IEEE Transactions on Pattern Analysis and Machine Intelligence, 23(7):689–706, july 2001.Google Scholar
- 8.C. A. Glasbey. An analysis of histogram-based thresholding algorithms. Computer Vision, Graphics, and Image Processing. Graphical Models and Image Processing, 55(6):532–537, November 1993.Google Scholar
- 11.N. Papamarkos and B. Gatos. A new approach for multilevel threshold selection. CVGIP: Graphical Models and Image Processing, 56(5):357–370, September 1994.Google Scholar
- 12.R. O. Duda, P. E. Hart, and D. G. Stork. Pattern Classification. John Wiley and Sons, 2001.Google Scholar