Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images

Abstract In view of the polysemy of mural images and the style difference among mural images painted in different dynasties as well as the high energy costs of the traditional manual dynasty classification method, which resorts to mural texts and historical documents, this study proposed an adaptive...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Jianfang Cao, Minmin Yan, Huiming Chen, Xiaodong Tian, Shang Ma
Formato:	article
Lenguaje:	EN
Publicado:	SpringerOpen 2021
Materias:	Preconvolution Even layer activation Adaptive enhancement Capsule network Mural dynasty classification Fine Arts N Analytical chemistry QD71-142
Acceso en línea:	https://doaj.org/article/bf01556c939849cd8f59f7df1cb6d37e
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:bf01556c939849cd8f59f7df1cb6d37e
record_format	dspace
spelling	oai:doaj.org-article:bf01556c939849cd8f59f7df1cb6d37e2021-11-07T12:15:44ZDynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images10.1186/s40494-021-00614-02050-7445https://doaj.org/article/bf01556c939849cd8f59f7df1cb6d37e2021-11-01T00:00:00Zhttps://doi.org/10.1186/s40494-021-00614-0https://doaj.org/toc/2050-7445Abstract In view of the polysemy of mural images and the style difference among mural images painted in different dynasties as well as the high energy costs of the traditional manual dynasty classification method, which resorts to mural texts and historical documents, this study proposed an adaptive enhancement capsule network (AECN) for automatic dynasty identification of mural images. Based on the original capsule network, we introduced a preconvolution structure to extract the high-level features of the mural images from Mogao Grottoes, such as color and texture. Then, we added an even activation operation to the layers of the network to enhance the fitting performance of the model. Finally, we performed adaptive modifications on the capsule network to increase the gradient smoothness of the model, based on which to optimize the model and thus to increase its classification precision. With the self-constructed DH1926 data set as the study subject, the proposed model achieved an accuracy of 84.44%, an average precision of 82.36%, an average recall rate of 83.75% and a comprehensive assessment score F1 of 83.96%. Compared with modified convolution neural networks and the original capsule network, the model proposed in study increased all the considered indices by more than 3%. It has a satisfactory fitting performance, which can extract the rich features of mural images at multiple levels and well express their semantic information. Furthermore, it has a higher accuracy and better robustness in the classification of the Mogao Grottoes murals, and therefore is of certain application values and research significance.Jianfang CaoMinmin YanHuiming ChenXiaodong TianShang MaSpringerOpenarticlePreconvolutionEven layer activationAdaptive enhancementCapsule networkMural dynasty classificationFine ArtsNAnalytical chemistryQD71-142ENHeritage Science, Vol 9, Iss 1, Pp 1-15 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Preconvolution Even layer activation Adaptive enhancement Capsule network Mural dynasty classification Fine Arts N Analytical chemistry QD71-142
spellingShingle	Preconvolution Even layer activation Adaptive enhancement Capsule network Mural dynasty classification Fine Arts N Analytical chemistry QD71-142 Jianfang Cao Minmin Yan Huiming Chen Xiaodong Tian Shang Ma Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
description	Abstract In view of the polysemy of mural images and the style difference among mural images painted in different dynasties as well as the high energy costs of the traditional manual dynasty classification method, which resorts to mural texts and historical documents, this study proposed an adaptive enhancement capsule network (AECN) for automatic dynasty identification of mural images. Based on the original capsule network, we introduced a preconvolution structure to extract the high-level features of the mural images from Mogao Grottoes, such as color and texture. Then, we added an even activation operation to the layers of the network to enhance the fitting performance of the model. Finally, we performed adaptive modifications on the capsule network to increase the gradient smoothness of the model, based on which to optimize the model and thus to increase its classification precision. With the self-constructed DH1926 data set as the study subject, the proposed model achieved an accuracy of 84.44%, an average precision of 82.36%, an average recall rate of 83.75% and a comprehensive assessment score F1 of 83.96%. Compared with modified convolution neural networks and the original capsule network, the model proposed in study increased all the considered indices by more than 3%. It has a satisfactory fitting performance, which can extract the rich features of mural images at multiple levels and well express their semantic information. Furthermore, it has a higher accuracy and better robustness in the classification of the Mogao Grottoes murals, and therefore is of certain application values and research significance.
format	article
author	Jianfang Cao Minmin Yan Huiming Chen Xiaodong Tian Shang Ma
author_facet	Jianfang Cao Minmin Yan Huiming Chen Xiaodong Tian Shang Ma
author_sort	Jianfang Cao
title	Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
title_short	Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
title_full	Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
title_fullStr	Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
title_full_unstemmed	Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
title_sort	dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images
publisher	SpringerOpen
publishDate	2021
url	https://doaj.org/article/bf01556c939849cd8f59f7df1cb6d37e
work_keys_str_mv	AT jianfangcao dynastyrecognitionalgorithmofanadaptiveenhancementcapsulenetworkforancientmuralimages AT minminyan dynastyrecognitionalgorithmofanadaptiveenhancementcapsulenetworkforancientmuralimages AT huimingchen dynastyrecognitionalgorithmofanadaptiveenhancementcapsulenetworkforancientmuralimages AT xiaodongtian dynastyrecognitionalgorithmofanadaptiveenhancementcapsulenetworkforancientmuralimages AT shangma dynastyrecognitionalgorithmofanadaptiveenhancementcapsulenetworkforancientmuralimages
_version_	1718443512890916864

Dynasty recognition algorithm of an adaptive enhancement capsule network for ancient mural images

Ejemplares similares