R Theta Local Neighborhood Pattern
R-Theta Local Neighborhood Pattern (RTLNP) is a local image descriptor used primarily in facial recognition and retrieval, aiming to improve accuracy and efficiency by encoding relationships between pixels within a local neighborhood using radial and angular information. Current research focuses on optimizing the encoding scheme to handle variations in pose, illumination, and other challenging conditions, often comparing RTLNP's performance against established methods like LBP and its variants. The development of robust and efficient local descriptors like RTLNP has significant implications for various image processing applications, particularly those dealing with unconstrained or noisy data.
Papers
Local Directional Gradient Pattern: A Local Descriptor for Face Recognition
Soumendu Chakraborty, Satish Kumar Singh, Pavan Chakraborty
Local Quadruple Pattern: A Novel Descriptor for Facial Image Recognition and Retrieval
Soumendu Chakraborty, Satish Kumar Singh, Pavan Chakraborty
R-Theta Local Neighborhood Pattern for Unconstrained Facial Image Recognition and Retrieval
Soumendu Chakraborty, Satish Kumar Singh, Pavan Chakraborty