A METHOD FOR AUTOMATICALLY DETECTING PEAKS IN THE GRAY-LEVEL HISTOGRAM OF A DIGITAL IMAGE

摘要

in many methods of digital image processing, it is desirable to apply selectively the treatment of recognizable structures in the image. it is known to select thresholds of grey level between the structures on the basis of the location of the corresponding peaks in the histogram of the grey level of the digital image.however, it is not easy to automatically detect the locations of the peaks in the histogram, and select the threshold of grey level.the present invention describes a method to automatically detect and identify the peaks of grey level thresholds for segmenting a digital image into separate structures.comprising the steps of detecting peaks in the histogram of the grey level of the digital image by using histogram equalization and the differentiation operator to generate a peak detection function, in which the zero crossings (di) of the posit if the negative of the function represents the beginning of a ridge.and the maximum values (bi) of such a zero crossing is the end of a ridge.