Resumo

ALVAREZ NAVARRO, Alejandro; MARANON REYES, Enrique Juan  e  OROZCO MORALES, Rubén. Review of facial recognition methods in RGB-D images acquired using a Kinect sensor. RCCI [online]. 2022, vol.16, n.2, pp. 157-187.  Epub 01-Jun-2022. ISSN 2227-1899.

Facial recognition in an image or video is a topic that has been researched and developed in recent years. However, it continues to be an open task with great complexities, from the scientific and economic point of view. Research is being carried out to develop or improve techniques and algorithms that are more effective than the existing ones and have lower computational costs. Three approaches are proposed, those based exclusively on RGB, depth images, or the fusion of both modalities. The objective of this work is offer a panoramic view of the proposals of face recognition using a Kinect sensor to acquire the images, realize a critical and comparative analysis of the main proposals and analyze the results obtained. From this analysis we will see the main contributions and evaluate the main research problems and challenges to be solved. Several descriptors for feature extraction and a set of classifiers are proposed in the articles reviewed in the period 2012-2021.

Palavras-chave : Facial recognition; Kinect; descriptors; feature extraction; classifiers.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )