激光与光电子学进展, 2020, 57 (8): 081105, 网络出版: 2020-04-03
基于多特征的彩色唐卡修复图像无参考质量评价方法 下载: 961次
No-Reference Quality Assessment Method for Inpainting Thangka Image Based on Multiple Features
成像系统 图像质量评价 高斯差分算子 色彩熵 AdaBoost神经网络 唐卡图像 imaging systems image quality assessment Gaussian difference operator color entropy AdaBoost neural network Thangka image
摘要
为了解决单一特征反映不同修复方法效果差异的局限性,结合唐卡图像结构特征和色彩特征,提出一种基于多特征的无参考彩色唐卡修复图像质量评价方法。首先,该算法利用唐卡图像纹理丰富的特性,选取高斯差分算子提取图像线描图,并结合唐卡对称特性获取图像结构特征;然后依据简单线性迭代聚类分割后不同超像素区域间的色彩熵变化差异,提取唐卡图像色彩特征;最后考虑到多尺度特征更加符合人眼视觉特性,将拉普拉斯金字塔分解后的多元特征输入自适应增强神经网络进行训练,预测得出修复图像质量客观评价分数。实验结果表明,该方法利用唐卡图像丰富的结构和色彩特征,较好地获得与主观评价一致的分数,其斯皮尔曼相关系数与皮尔逊相关系数均达到0.94以上。
Abstract
This paper proposes a new no-reference quality assessment method for inpainting Thangka image based on multiple features, and combining the structural and color characteristic of Thangka images to solve the problem that a single feature is confined to reflects the difference in the effect of restoration methods. The proposed algorithm not only uses the rich texture of Thangka images but also selects Gaussian difference operator to extract the line drawing of target image, and combining symmetry characteristic of Thangka image to obtain structural features. Secondly, the color features of Thangka images are extracted according to the difference of color entropy between each superpixels after simple linear iterative cluster segmentation. Finally, considering that the multi-scale features are more consistent with the human visual characteristics, the decomposed image features are input into the adaptive neural network for training, and the objective evaluation score of image quality is predicted. The experimental results show that this method can obtain the scores which is consistent with the subjective evaluation by utilizing the structure and color characteristics of Thangka images, and its Spearman correlation coefficient and Pearson correlation coefficient are both above 0.94.
叶雨琪, 胡文瑾. 基于多特征的彩色唐卡修复图像无参考质量评价方法[J]. 激光与光电子学进展, 2020, 57(8): 081105. Yuqi Ye, Wenjin Hu. No-Reference Quality Assessment Method for Inpainting Thangka Image Based on Multiple Features[J]. Laser & Optoelectronics Progress, 2020, 57(8): 081105.