Perceptual image quality assessment(IQA)is one of the most indispensable yet challenging problems in image processing and computer vision.It is quite necessary to develop automatic and efficient approaches that can ac...Perceptual image quality assessment(IQA)is one of the most indispensable yet challenging problems in image processing and computer vision.It is quite necessary to develop automatic and efficient approaches that can accurately predict perceptual image quality consistently with human subjective evaluation.To further improve the prediction accuracy for the distortion of color images,in this paper,we propose a novel effective and efficient IQA model,called perceptual gradient similarity deviation(PGSD).Based on the gradient magnitude similarity,we proposed a gradient direction selection method to automatically determine the pixel-wise perceptual gradient.The luminance and chrominance channels are both took into account to characterize the quality degradation caused by intensity and color distortions.Finally,a multi-scale strategy is utilized and pooled with different weights to incorporate image details at different resolutions.Experimental results on LIVE,CSIQ and TID2013 databases demonstrate the superior performances of the proposed algorithm.展开更多
A semi-reference image quality assessment metric based on similarity measurement for synthesized virtual viewpoint image （VVI） in free-viewpoint television system （FFV） is proposed in this paper. The key point of ...A semi-reference image quality assessment metric based on similarity measurement for synthesized virtual viewpoint image （VVI） in free-viewpoint television system （FFV） is proposed in this paper. The key point of the proposed metric is taking resemblant information between VVI and its neighbor view images for quality assessment to make our metric to be extended to multi-semi-reference image quality assessment easily. The proposed metric first extracts impact factors from image features, then combines an image synthesis technique and similarity functions, in which, disparity information are taken into account for registering the resemblant regions. Experiments are divided into three phases. Phase I is to verify the validation of the proposed metric by taking impaired images and original reference into account. The experimental results show the agreement between evaluation scores and bio-characteristic of human visual system. Phase II shows the accordance with Phase I by taking neighbor view as reference. The proposed metric can be taken as a full reference one to evaluate the image quality even though the original reference is absent. Phase III is then performed to evaluate the quality of WI. Evaluation scores in the experimental results are able to evaluate the quality of VVI.展开更多
To evaluate the quality of blurred images effectively,this study proposes a no-reference blur assessment method based on gradient distortion measurement and salient region maps.First,a Gaussian low-pass filter is used...To evaluate the quality of blurred images effectively,this study proposes a no-reference blur assessment method based on gradient distortion measurement and salient region maps.First,a Gaussian low-pass filter is used to construct a reference image by blurring a given image.Gradient similarity is included to obtain the gradient distortion measurement map,which can finely reflect the smallest possible changes in textures and details.Second,a saliency model is utilized to calculate image saliency.Specifically,an adaptive method is used to calculate the specific salient threshold of the blurred image,and the blurred image is binarized to yield the salient region map.Block-wise visual saliency serves as the weight to obtain the final image quality.Experimental results based on the image and video engineering database,categorial image quality database,and camera image database demonstrate that the proposed method correlates well with human judgment.Its computational complexity is also relatively low.展开更多
Objective image quality measure, which is a fundamental and challenging job in image processing, evaluates the image quality consistently with human perception automatically. On the assumption that any image distortio...Objective image quality measure, which is a fundamental and challenging job in image processing, evaluates the image quality consistently with human perception automatically. On the assumption that any image distortion could be modeled as the difference between the directional projection-based maps of reference and distortion images, we propose a new objective quality assessment method based on directional projection for full reference model. Experimental results show that the proposed metrics are well consistent with the subjective quality score.展开更多
In this paper, a saliency weighted visual feature similarity （SWVFS） metric is proposed for full reference im- age quality assessment （IQA）. Instead of traditional spatial pooling strategies, a visual saliency-bas...In this paper, a saliency weighted visual feature similarity （SWVFS） metric is proposed for full reference im- age quality assessment （IQA）. Instead of traditional spatial pooling strategies, a visual saliency-based approach is em- ployed for better compliance with properties of the human visual system, where the saliency allocation is closely related to the activity of posterior parietal cortex and the pluvial nu- clei of the thalamus. Assuming that the saliency map actually represents the contribution of locally computed visual distor- tions to the overall image quality, the gradient similarity and the textural congruency are merged into the final image qual- ity indicator. The gradient and texture comparison play com- plementary roles in characterizing the local image distortion. Extensive experiments conducted on seven publicly available image databases show that the performance of SWVFS is competitive with the state-of-the-art IQA algorithms.展开更多
Recent developments in 3D graphics technology have led to extensive processes on 3D meshes(e.g.,compression,simplification,transmission and watermarking),these processes unavoidably cause the visual perceptual degrada...Recent developments in 3D graphics technology have led to extensive processes on 3D meshes(e.g.,compression,simplification,transmission and watermarking),these processes unavoidably cause the visual perceptual degradation of the 3D objects.The existing mesh visual quality evaluation metrics either require topology constrain or fail to reflect the perceived visual quality.Meanwhile,for the 3D objects that are observed on 2D screens by the users,it is reasonable to apply image metric to assess the distortion caused by mesh simplification.We attempt to explore the efficiency of image metric for assessing the visual fidelity of the simplified 3D model in this paper.For this purpose,several latest and most effective image metrics,2D snapshots,number and pooling algorithms are involved in our study,and finally tested on the IEETA simplification database.The statistical data allow the researcher to select the optimal parameter for this image-based mesh visual quality assessment and provide a new perspective for the design and performance assessment of mesh simplification algorithms.展开更多
在SSIM算法的基础上,结合亮度和对比度掩蔽等视觉感知信息构造视觉感知(visual perception)函数,提出基于视觉感知的梯度结构相似度评价方法VI_GSSIM(visual perception and gradient based SSIM)。该方法通过图像质量与图像内容和失真...在SSIM算法的基础上,结合亮度和对比度掩蔽等视觉感知信息构造视觉感知(visual perception)函数,提出基于视觉感知的梯度结构相似度评价方法VI_GSSIM(visual perception and gradient based SSIM)。该方法通过图像质量与图像内容和失真类型的相关性,结合图像的误差可视性与内容可视性构造视觉感知函数,对HVS底层视觉系统建模;同时利用梯度重新定义结构信息,得到基于视觉感知的梯度结构相似度模型,对图像进行质量评价。实验结果表明,提出的VI_GSSIM算法比SSIM更符合人眼的视觉特性,尤其适合评价降质较严重的图像。展开更多
文摘Perceptual image quality assessment(IQA)is one of the most indispensable yet challenging problems in image processing and computer vision.It is quite necessary to develop automatic and efficient approaches that can accurately predict perceptual image quality consistently with human subjective evaluation.To further improve the prediction accuracy for the distortion of color images,in this paper,we propose a novel effective and efficient IQA model,called perceptual gradient similarity deviation(PGSD).Based on the gradient magnitude similarity,we proposed a gradient direction selection method to automatically determine the pixel-wise perceptual gradient.The luminance and chrominance channels are both took into account to characterize the quality degradation caused by intensity and color distortions.Finally,a multi-scale strategy is utilized and pooled with different weights to incorporate image details at different resolutions.Experimental results on LIVE,CSIQ and TID2013 databases demonstrate the superior performances of the proposed algorithm.
基金Supported by the National Natural Science Foundation of China (No. 60672073,60872094)the Program for New Century Excellent Talents in University (NCET-06-0537)the Natural Science Foundation of Ningbo (No. 2007A610037).
文摘A semi-reference image quality assessment metric based on similarity measurement for synthesized virtual viewpoint image （VVI） in free-viewpoint television system （FFV） is proposed in this paper. The key point of the proposed metric is taking resemblant information between VVI and its neighbor view images for quality assessment to make our metric to be extended to multi-semi-reference image quality assessment easily. The proposed metric first extracts impact factors from image features, then combines an image synthesis technique and similarity functions, in which, disparity information are taken into account for registering the resemblant regions. Experiments are divided into three phases. Phase I is to verify the validation of the proposed metric by taking impaired images and original reference into account. The experimental results show the agreement between evaluation scores and bio-characteristic of human visual system. Phase II shows the accordance with Phase I by taking neighbor view as reference. The proposed metric can be taken as a full reference one to evaluate the image quality even though the original reference is absent. Phase III is then performed to evaluate the quality of WI. Evaluation scores in the experimental results are able to evaluate the quality of VVI.
基金The National Natural Science Foundation of China(No.61762004,61762005)the National Key Research and Development Program(No.2018YFB1702700)+1 种基金the Science and Technology Project Founded by the Education Department of Jiangxi Province,China(No.GJJ200702,GJJ200746)the Open Fund Project of Jiangxi Engineering Laboratory on Radioactive Geoscience and Big Data Technology(No.JETRCNGDSS201901,JELRGBDT202001,JELRGBDT202003).
文摘To evaluate the quality of blurred images effectively,this study proposes a no-reference blur assessment method based on gradient distortion measurement and salient region maps.First,a Gaussian low-pass filter is used to construct a reference image by blurring a given image.Gradient similarity is included to obtain the gradient distortion measurement map,which can finely reflect the smallest possible changes in textures and details.Second,a saliency model is utilized to calculate image saliency.Specifically,an adaptive method is used to calculate the specific salient threshold of the blurred image,and the blurred image is binarized to yield the salient region map.Block-wise visual saliency serves as the weight to obtain the final image quality.Experimental results based on the image and video engineering database,categorial image quality database,and camera image database demonstrate that the proposed method correlates well with human judgment.Its computational complexity is also relatively low.
文摘Objective image quality measure, which is a fundamental and challenging job in image processing, evaluates the image quality consistently with human perception automatically. On the assumption that any image distortion could be modeled as the difference between the directional projection-based maps of reference and distortion images, we propose a new objective quality assessment method based on directional projection for full reference model. Experimental results show that the proposed metrics are well consistent with the subjective quality score.
文摘In this paper, a saliency weighted visual feature similarity （SWVFS） metric is proposed for full reference im- age quality assessment （IQA）. Instead of traditional spatial pooling strategies, a visual saliency-based approach is em- ployed for better compliance with properties of the human visual system, where the saliency allocation is closely related to the activity of posterior parietal cortex and the pluvial nu- clei of the thalamus. Assuming that the saliency map actually represents the contribution of locally computed visual distor- tions to the overall image quality, the gradient similarity and the textural congruency are merged into the final image qual- ity indicator. The gradient and texture comparison play com- plementary roles in characterizing the local image distortion. Extensive experiments conducted on seven publicly available image databases show that the performance of SWVFS is competitive with the state-of-the-art IQA algorithms.
文摘Recent developments in 3D graphics technology have led to extensive processes on 3D meshes(e.g.,compression,simplification,transmission and watermarking),these processes unavoidably cause the visual perceptual degradation of the 3D objects.The existing mesh visual quality evaluation metrics either require topology constrain or fail to reflect the perceived visual quality.Meanwhile,for the 3D objects that are observed on 2D screens by the users,it is reasonable to apply image metric to assess the distortion caused by mesh simplification.We attempt to explore the efficiency of image metric for assessing the visual fidelity of the simplified 3D model in this paper.For this purpose,several latest and most effective image metrics,2D snapshots,number and pooling algorithms are involved in our study,and finally tested on the IEETA simplification database.The statistical data allow the researcher to select the optimal parameter for this image-based mesh visual quality assessment and provide a new perspective for the design and performance assessment of mesh simplification algorithms.
文摘在SSIM算法的基础上,结合亮度和对比度掩蔽等视觉感知信息构造视觉感知(visual perception)函数,提出基于视觉感知的梯度结构相似度评价方法VI_GSSIM(visual perception and gradient based SSIM)。该方法通过图像质量与图像内容和失真类型的相关性,结合图像的误差可视性与内容可视性构造视觉感知函数,对HVS底层视觉系统建模;同时利用梯度重新定义结构信息,得到基于视觉感知的梯度结构相似度模型,对图像进行质量评价。实验结果表明,提出的VI_GSSIM算法比SSIM更符合人眼的视觉特性,尤其适合评价降质较严重的图像。
