# What is the difference between HOG and LBP?

Source: https://zilliz.com/ai-faq/what-is-the-difference-between-hog-and-lbp
Author: zilliz

HOG (Histogram of Oriented Gradients) and LBP (Local Binary Patterns) are feature extraction techniques used in image processing, but they focus on different aspects of an image. HOG emphasizes the gradient and edge direction, while LBP focuses on local texture patterns. HOG computes the gradient orientation in an image and creates a histogram of these directions within small regions. It is commonly used for tasks like object detection, especially pedestrian detection, due to its ability to capture edge structure effectively. LBP, on the other hand, examines the relationship between a pixel and its surrounding neighbors, encoding these patterns into binary representations. It is widely used in texture classification and facial recognition. HOG works better for shape-based tasks, while LBP is suited for texture-based analysis.