Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance
In this paper a robust algorithm against illumination changes for skin detection in images is proposed. A database with 50 controlled condition images and 50 without controlled conditions of people in frontal position showing face, hands and arms was used. Five algorithms to perform color correction...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | info:eu-repo/semantics/publishedVersion |
| Language: | spa |
| Published: |
Universidad Antonio Nariño
2021
|
| Subjects: | |
| Online Access: | http://revistas.uan.edu.co/index.php/ingeuan/article/view/389 http://repositorio.uan.edu.co/handle/123456789/3952 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1812647829528641536 |
|---|---|
| author | Arévalo Casallas, Diego Armando Castañeda Obando, David Ricardo Castañeda Fandiño, Jos´é Ignacio |
| author_facet | Arévalo Casallas, Diego Armando Castañeda Obando, David Ricardo Castañeda Fandiño, Jos´é Ignacio |
| author_sort | Arévalo Casallas, Diego Armando |
| collection | DSpace |
| description | In this paper a robust algorithm against illumination changes for skin detection in images is proposed. A database with 50 controlled condition images and 50 without controlled conditions of people in frontal position showing face, hands and arms was used. Five algorithms to perform color correction are evaluated: Simple Correction with Green Channel, Color Channel Compression, Color Channel Expansion, Fixed Reference and Gamma Correction. And four algorithms for segmentation are evaluated as well: RGB Skin Color, Reference Histogram, Euclidean Distance and Mahalanobis Distance. The proposed algorithm uses the Fixed Reference method together with Gamma Correction for color correction and performs the skin segmentation based on an RCrR color plane, found by making the transformation of the images using RGB and YCbCr spaces, finally Mahalanobis Distance is used. An average sensitivity value of 99.36 % and specificity of 84.31 % were obtained as result. |
| format | info:eu-repo/semantics/publishedVersion |
| id | repositorio.uan.edu.co-123456789-3952 |
| institution | Repositorio Digital UAN |
| language | spa |
| publishDate | 2021 |
| publisher | Universidad Antonio Nariño |
| record_format | dspace |
| spelling | repositorio.uan.edu.co-123456789-39522024-10-09T23:03:10Z Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance Segmentación Visual Robusta utilizando el Plano RCrR y la Distancia de Mahalanobis Arévalo Casallas, Diego Armando Castañeda Obando, David Ricardo Castañeda Fandiño, Jos´é Ignacio Faded photo correction gray world assumption gamma correction illumination skin color segmentation euclidean distance mahalanobis distance histogram Corrección foto descolorida Suposición de mundo gris Corrección gamma iluminación segmentación color de piel Distancia Euclidiana Distancia Mahalanobis Histograma In this paper a robust algorithm against illumination changes for skin detection in images is proposed. A database with 50 controlled condition images and 50 without controlled conditions of people in frontal position showing face, hands and arms was used. Five algorithms to perform color correction are evaluated: Simple Correction with Green Channel, Color Channel Compression, Color Channel Expansion, Fixed Reference and Gamma Correction. And four algorithms for segmentation are evaluated as well: RGB Skin Color, Reference Histogram, Euclidean Distance and Mahalanobis Distance. The proposed algorithm uses the Fixed Reference method together with Gamma Correction for color correction and performs the skin segmentation based on an RCrR color plane, found by making the transformation of the images using RGB and YCbCr spaces, finally Mahalanobis Distance is used. An average sensitivity value of 99.36 % and specificity of 84.31 % were obtained as result. En este artículo se propone un algoritmo robusto ante los cambios de iluminación para la detección de la piel en imágenes, se utiliza una base de datos que consta de 50 imágenes en condiciones controladas y 50 en condiciones no controladas, las imágenes cuentan con personas en forma frontal, mostrando rostro, manos, y brazos. Se evalúan 5 algoritmos para realizar corrección de color los cuales son: Corrección sencilla con canal verde, Compresión canal de color, Expansión canal de color, Referencia fija, Corrección Gamma. Se evalúan 4 algoritmos para segmentación los cuales son: Color de piel en RGB, Referencia de Histograma, Distancia Euclidiana y Distancia de Mahalanobis. El algoritmo propuesto utiliza el método referencia fija unido al algoritmo de corrección gamma para corrección de color y realiza segmentación de la piel a partir de un plano de color RCrR, encontrado de la transformación de las imágenes utilizando los espacios RGB y YCbCr, finalmente utiliza la distancia de Mahalanobis. Como resultado se obtiene un valor promedio de sensibilidad igual 99.36% y de especificidad igual 84.31%. 2021-06-16T13:53:12Z 2021-06-16T13:53:12Z 2014-12-20 info:eu-repo/semantics/publishedVersion http://purl.org/coar/resource_type/c_6501 info:eu-repo/semantics/article http://purl.org/coar/version/c_970fb48d4fbd8a85 http://revistas.uan.edu.co/index.php/ingeuan/article/view/389 http://repositorio.uan.edu.co/handle/123456789/3952 spa http://revistas.uan.edu.co/index.php/ingeuan/article/view/389/328 Acceso abierto Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) https://creativecommons.org/licenses/by-nc-sa/4.0/ info:eu-repo/semantics/openAccess http://purl.org/coar/access_right/c_abf2 application/pdf Universidad Antonio Nariño 2346-1446 2145-0935 INGE@UAN - TENDENCIAS EN LA INGENIERÍA; Vol. 5 Núm. 9 (2014) |
| spellingShingle | Faded photo correction gray world assumption gamma correction illumination skin color segmentation euclidean distance mahalanobis distance histogram Corrección foto descolorida Suposición de mundo gris Corrección gamma iluminación segmentación color de piel Distancia Euclidiana Distancia Mahalanobis Histograma Arévalo Casallas, Diego Armando Castañeda Obando, David Ricardo Castañeda Fandiño, Jos´é Ignacio Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title | Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title_full | Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title_fullStr | Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title_full_unstemmed | Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title_short | Robust Visual Segmentation using RCrR Plane and Mahalanobis Distance |
| title_sort | robust visual segmentation using rcrr plane and mahalanobis distance |
| topic | Faded photo correction gray world assumption gamma correction illumination skin color segmentation euclidean distance mahalanobis distance histogram Corrección foto descolorida Suposición de mundo gris Corrección gamma iluminación segmentación color de piel Distancia Euclidiana Distancia Mahalanobis Histograma |
| url | http://revistas.uan.edu.co/index.php/ingeuan/article/view/389 http://repositorio.uan.edu.co/handle/123456789/3952 |
| work_keys_str_mv | AT arevalocasallasdiegoarmando robustvisualsegmentationusingrcrrplaneandmahalanobisdistance AT castanedaobandodavidricardo robustvisualsegmentationusingrcrrplaneandmahalanobisdistance AT castanedafandinojoseignacio robustvisualsegmentationusingrcrrplaneandmahalanobisdistance AT arevalocasallasdiegoarmando segmentacionvisualrobustautilizandoelplanorcrryladistanciademahalanobis AT castanedaobandodavidricardo segmentacionvisualrobustautilizandoelplanorcrryladistanciademahalanobis AT castanedafandinojoseignacio segmentacionvisualrobustautilizandoelplanorcrryladistanciademahalanobis |
