Image histogram

• It is mainly used to describe the frequency distribution of pixel brightness or color in an image.
• It is a statistical graph where the horizontal axis represents the range of pixel values (gray levels or color intensities), and the vertical axis represents the number of pixels that have each value.
• Histograms don’t encode information about the spatial arrangement of pixels in the image

• Binning: Grouping continuous or high-resolution values into discrete intervals (bins) for statistical counting without changing the original data.
• Quantization: Mapping continuous or high-precision values to a smaller set of discrete levels, altering the data itself.

Properties

• Contrast: the range of intensity values effectively used within an image
• Dynamic Range: the range (or number) of distinct intensity values that can be represented in an image or system.

Image Defects

• Saturation: the illumination values lying outside of the sensor’s range are mapped to its maximum or minimum values: spike at the tails
  • Intensity saturation: pixels concentrated near 0 or 255
• Exposure
  • Under-exposure: Most pixel intensities cluster near 0, making the image too dark with loss of shadow detail.
  • Over-exposure: Most pixel intensities cluster near 255, making the image too bright with loss of highlight detail.

Color Image Histogram

Two kinds of histogram:

• Intensity histogram
• Individual Color Channel Histograms

Cumulative Histograms

• K: The total number of possible gray levels in the image (for an 8-bit image, K = 256)
• M × N: The total number of pixels in the image