武汉理工大学数字图像处理英文课件 Chapter3（9）.pptVIP

Gray-scale Histograms Histogram as Area Function Image As 2-D Random Field Higher Dimensional Histograms Histogram Properties Histogram Equalization Histogram Uses * Number of Pixels at Each Grey-level or in Range of Grey Levels Plot of Frequency of Occurrence as Function of Pixel Value Probability Density Function (pdf) Pixel Value Number of Occurrances 0 255 Area Function of Continuous Image Number of Pixels at One Grey-Level can be Considered to Represent an Area Area function proportional to Cumulative Distribution Function Histogram proportional to Probability Density Function Color Images 2-D Histogram, One Color on Each Axis Joint pdf Marginal pdf on each axis Use to develop discriminant function Histogram is a Many-to-One Mapping Different Images May Have Same Histogram Histogram Is Invariant Under Certain Image Operations If Object has Constant Grey-Level or Range of Grey Levels, then Relative Area is Sum of Histogram Lines at that Level(s) Normalizing Histogram to 1 Makes it a pdf Histogram of Summed Image is Sum of Histograms of Individual Images. A Mapping of Grey Levels”P” Into Grey Levels “q” So That the Distribution of Grey Levels in “q” Is (Approx.) Uniform Stretching Is NOT Histogram Equalization Based on Viewing the Histogram As a pdf and Integrating to Get the Desired Cumulative Density Function *