Local Binary Pattern (LBP)

Task: Converting multiple images in a folder (also a single image) to LBP images

Local-Binary-Pattern

Local binary patterns are a features used for classification in the field of machine vision. It is a powerful powerful feature in texture classification; if the local binary pattern feature is combined with the directional gradient histogram, it can effectively improve the detection effect. Local binary mode is a simple but very efficient texture operator. It compares each pixel with its neighbors and saves the result as a binary number. Due to its strong discriminative power and computational simplicity, local binary pattern texture operators have been applied in different scenarios. The most important property of LBP is robustness to grayscale changes such as illumination changes. Another important feature of it is its computational simplicity, which enables real-time analysis of images.

LBP image calculation

This is the code for preproessing original images to local binary pattern (LBP) images. This code will help researcher process their images into LBP. This code process single and multiple images in a folder and save them in another folder. ensure the right directory path is called properly. The result of the code is presented below:

Original CT image

LBP CT image