Grid Color Movement In Image Analysis

January 29, 2018 Author: munishmishra04_3od47tgp
Print Friendly, PDF & Email

Content-based image Retrieval techniques can be divided into two main domains: pixel and compressed domain techniques. In the pixel domain, the values of individual pixels in the image matrix are used directly for making visual indexes. In the compressed domain, transformed data, which is the result of mapping the original image matrix into another domain, is employed for feature extraction and retrieval. One of the main tasks for Content-based image Retrieval (CBIR) systems is similarity comparison, extracting feature of every image based on its pixel values and defining rules for comparing images. These features become the image representation for measuring similarity with other images in the database. Images are compared by calculating the difference of its feature components to other image descriptors i.e namely colour, texture and shape features. In this article we are going to discuss the color descriptors.

Grid Color Movement


Colour feature is one of the most widely used features in low level feature. Associated with shape feature, Colour feature and texture feature shows better stability and is more insensitive to the rotation and zoom of the image. Colour not only adds beauty to objects but also more information that is used as a powerful tool in content-based image retrieval. In Colour index, given a query image, the objective is to retrieve all the images whose Colour and texture compositions are similar to those of the query image. In Colour image retrieval there are various methods, but here we will discuss some prominent methods. The feature vector we will use is called “Grid-based Colour
Moment”. To compute this feature vectors for a given image: [1]

  • Convert the image from RGB for HSV Colour space (Hint: use the function rgb2hsv in Matlab for this operation)
  • Regularly divide the image into 3×3 blocks
  • For each of these nine blocks
  • Compute its mean Colour (H/S/V)
  • \( x’=1/N ∑_{i=1}^Nx_i \)

Where the number of pixels within each block is given by N, and ​\( x_i \)​ is the pixel intensity in H/S/V channels.

  • Compute its variance (H/S/V)
  • \[ σ^2=1/N∑_{(i=1)}^N(x_i-x’)^2 \]

    Compute its skewness (H/S/V)

\[ γ= (1/n ∑_{(i=1)}^N(x_i-x’)^3 )/(1/n ∑_{(i=1)}^N(x_i-x’)^2 )^{3⁄2} \]

  • Each block will have 3+3+3=9 features, and thus the entire image will have 9×9=81 features. Before we use SVM to train the classifier, we first need to normalize the 81 features to be within the same range, in order to achieve good numerical behavior. To do the normalization, for each of the 81 features:
  • Compute the mean and standard deviation from the training dataset

\[ μ= 1/M ∑_{(i=1)}^Mf_i \]

\[ σ=√(1/M ∑_{(i=0)}^M(f_i-μ)^2 ) \]

  • where M is the number of images in the training dataset, and f_i is the feature of the i-th training sample.
  • Perform the “whitening” transform for all the data (including both the training data and the testing data), and get the normalized feature value:

\[ f’_i=(f_i-μ)/σ \]


[1] Atul Nandwal, “An Efficient Face Recognition Algorithm Using Colour Grid Movement Analysis”, International Journal of Research in Computer Engineering and Electronics, Page # 1 ISSN 2319-376X VOl : 2 ISSUE :6

[2] Juan M. Bands, Rafal A. Angryk, “Selection of Image Parameters as the First Step Towards Creating a CBIR System for the Solar Dynamics Observatory”, Montana State University Bozeman, MT USA

No Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

Insert math as
Additional settings
Formula color
Text color
Type math using LaTeX
Nothing to preview