/siv-texture-analysis

Implementation of Autocorrelation Function, Local Binary Patterns, and Co-Occurrence Matrices for images' texture analysis

Primary LanguagePythonMIT LicenseMIT

Signal, Image and Video - Analysis of Fabric Textures project

The objective of the project was to develop a small module to be included in a broader tool that could be used by future students in the laboratory lessons of the Signal, Image and Video course. This tool aims to permit a graphical building of elaboration chains for signals, using blocks with predefined inputs and outputs.

The topic identified for this module development was the analysis and synthesis of fabric textures, specifically on statistical models. The identified techniques of Autocorrelation Function, Local Binary Patterns and Co-Occurrence Matrices should have been used to extract their relative parameters as outputs from an input image. These parameters will be used by future works to identify defects or synthesize a texture.