Image Histogram

 Histogram

Color is one of the features that can be used in image retrieval system. Color can be represented in the form of a histogram. Color histogram represents the distribution of the number of pixels for each intensity colors in the image. To define the histogram, each color quantize into some discrete level. For each level the number of pixels is calculated according to the value  as shown in Figure below

Horizontal axis represent color levels of the pixels in image and vertical axis express pixel numbers in each appropriate color level.   

 Histogram Distance

Color composition similarity of images  can be determined by its color histogram. Level of similarity between two images measured with its histogram distance.  If G =  {g₁, g₂, . . . , g_N } and H =  {h₁, h₂, . . . , h_N } are the color histogram from two images, where g_i and h_i is the number of pixels at level i  and N is the number of levels for each histogram, then distance (d) between two histogram can be expressed in Manhattan metric as in equation :

Image with smaller distance value is considered has more similar color compared with image that has greater distance value.