7 Local Patterns for Feature Extraction
7.1 Quantized Neighborhood Local Intensity Extrema Patterns for Image Retrieval
7.1.1 Introduction
P. Banerjee et al. (2018) introduced the local intensity patterns. It is an extended concept of Local Binary Patterns (LBP). Relationship among pixels in a neighborhood is used in this method. Major limitations in terms of the neighborhood and increased feature vector length influenced us to propose Quantized neighborhood local intensity extrema patterns for texture extraction. Structures in four directions, i.e. horizontal, vertical, diagonal and anti-diagonal are taken. In other words, the image is quantized in terms of number of gray levels. We use a 5 × 5 window ...
Get Image Pattern Recognition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
