The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Une nouvelle approche de segmentation des superpixels pilotée par un modèle de mélange uniforme avec contraintes spatiales (UMMS) est proposée. Selon cet algorithme, chaque observation, c'est-à-dire chaque pixel, est d'abord représentée comme un vecteur à cinq dimensions composé de couleurs dans l'espace CLELAB et d'informations de position. Et puis, nous définissons une nouvelle distribution uniforme en ajoutant la position des pixels, afin que cette distribution puisse décrire chaque pixel de l'image d'entrée. Norme 1 pondérée appliquée à la différence entre les pixels et permettant de contrôler la compacité du superpixel. De plus, un schéma efficace d’estimation des paramètres est introduit pour réduire la complexité des calculs. Plus précisément, la probabilité a priori invariante et la plage de paramètres limitent la localisation des superpixels, et la technique robuste d'optimisation moyenne garantit la précision des limites des superpixels. Enfin, chaque distribution uniforme définie est associée à un superpixel et l'UMMS proposé implémente avec succès la segmentation des superpixels. Les expériences sur l'ensemble de données BSDS500 vérifient que l'UMMS surpasse la plupart des approches de pointe en termes de précision, de régularité et de rapidité de segmentation.
Pengyu WANG
East China University of Science and Technology
Hongqing ZHU
East China University of Science and Technology
Ning CHEN
East China University of Science and Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Pengyu WANG, Hongqing ZHU, Ning CHEN, "UMMS: Efficient Superpixel Segmentation Driven by a Mixture of Spatially Constrained Uniform Distribution" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 1, pp. 181-185, January 2020, doi: 10.1587/transinf.2019EDL8134.
Abstract: A novel superpixel segmentation approach driven by uniform mixture model with spatially constrained (UMMS) is proposed. Under this algorithm, each observation, i.e. pixel is first represented as a five-dimensional vector which consists of colour in CLELAB space and position information. And then, we define a new uniform distribution through adding pixel position, so that this distribution can describe each pixel in input image. Applied weighted 1-Norm to difference between pixels and mean to control the compactness of superpixel. In addition, an effective parameter estimation scheme is introduced to reduce computational complexity. Specifically, the invariant prior probability and parameter range restrict the locality of superpixels, and the robust mean optimization technique ensures the accuracy of superpixel boundaries. Finally, each defined uniform distribution is associated with a superpixel and the proposed UMMS successfully implements superpixel segmentation. The experiments on BSDS500 dataset verify that UMMS outperforms most of the state-of-the-art approaches in terms of segmentation accuracy, regularity, and rapidity.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019EDL8134/_p
Copier
@ARTICLE{e103-d_1_181,
author={Pengyu WANG, Hongqing ZHU, Ning CHEN, },
journal={IEICE TRANSACTIONS on Information},
title={UMMS: Efficient Superpixel Segmentation Driven by a Mixture of Spatially Constrained Uniform Distribution},
year={2020},
volume={E103-D},
number={1},
pages={181-185},
abstract={A novel superpixel segmentation approach driven by uniform mixture model with spatially constrained (UMMS) is proposed. Under this algorithm, each observation, i.e. pixel is first represented as a five-dimensional vector which consists of colour in CLELAB space and position information. And then, we define a new uniform distribution through adding pixel position, so that this distribution can describe each pixel in input image. Applied weighted 1-Norm to difference between pixels and mean to control the compactness of superpixel. In addition, an effective parameter estimation scheme is introduced to reduce computational complexity. Specifically, the invariant prior probability and parameter range restrict the locality of superpixels, and the robust mean optimization technique ensures the accuracy of superpixel boundaries. Finally, each defined uniform distribution is associated with a superpixel and the proposed UMMS successfully implements superpixel segmentation. The experiments on BSDS500 dataset verify that UMMS outperforms most of the state-of-the-art approaches in terms of segmentation accuracy, regularity, and rapidity.},
keywords={},
doi={10.1587/transinf.2019EDL8134},
ISSN={1745-1361},
month={January},}
Copier
TY - JOUR
TI - UMMS: Efficient Superpixel Segmentation Driven by a Mixture of Spatially Constrained Uniform Distribution
T2 - IEICE TRANSACTIONS on Information
SP - 181
EP - 185
AU - Pengyu WANG
AU - Hongqing ZHU
AU - Ning CHEN
PY - 2020
DO - 10.1587/transinf.2019EDL8134
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 1
JA - IEICE TRANSACTIONS on Information
Y1 - January 2020
AB - A novel superpixel segmentation approach driven by uniform mixture model with spatially constrained (UMMS) is proposed. Under this algorithm, each observation, i.e. pixel is first represented as a five-dimensional vector which consists of colour in CLELAB space and position information. And then, we define a new uniform distribution through adding pixel position, so that this distribution can describe each pixel in input image. Applied weighted 1-Norm to difference between pixels and mean to control the compactness of superpixel. In addition, an effective parameter estimation scheme is introduced to reduce computational complexity. Specifically, the invariant prior probability and parameter range restrict the locality of superpixels, and the robust mean optimization technique ensures the accuracy of superpixel boundaries. Finally, each defined uniform distribution is associated with a superpixel and the proposed UMMS successfully implements superpixel segmentation. The experiments on BSDS500 dataset verify that UMMS outperforms most of the state-of-the-art approaches in terms of segmentation accuracy, regularity, and rapidity.
ER -