Yuukou Horita

No reference image quality assessment for JPEG2000 based on spatial features

Perceptual image quality evaluation has become an important issue, due to increasing transmission of multimedia contents over the Internet and 3G mobile networks. Most of the no reference perceptual image quality evaluations traditionally attempted to quantify the predefined artifacts of the coded images. Under the assumption that human visual perception is very sensitive to edge information of an image and any kinds of artifacts create pixel distortion, we propose a new approach for designing a no reference image quality evaluation model for JPEG2000 images in this paper, which uses pixel distortions and edge information. Subjective experiment results on the images are used to train and test the model, which has achieved good quality prediction performance.

No-reference stereoscopic image quality assessment

Display of stereo images is widely used to enhance the viewing experience of three-dimensional imaging and communication systems. In this paper, we propose a method for estimating the quality of stereoscopic images using segmented image features and disparity. This method is inspired by the human visual system. We believe the perceived distortion and disparity of any stereoscopic display is strongly dependent on local features, such as edge (non-plane) and non-edge (plane) areas. Therefore, a no-reference perceptual quality assessment is developed for JPEG coded stereoscopic images based on segmented local features of artifacts and disparity. Local feature information such as edge and non-edge area based relative disparity estimation, as well as the blockiness and the blur within the block of images are evaluated in this method. Two subjective stereo image databases are used to evaluate the performance of our method. The subjective experiments results indicate our model has sufficient prediction performance.

Stereoscopic image quality prediction

Three-dimensional (3D) imaging has attracted considerable attention recently due to its increasingly wide range of applications. Consequently, perceived quality is a great important issue to assess the performance of all 3D imaging applications. Perceived distortion and depth of any stereoscopic images are strongly dependent on the local features, such as edge, flat and texture. In this paper, we propose an noreference (NR) perceptual quality assessment for JPEG coded stereoscopic images based on segmented local features of artifacts and disparity. The local features information of stereoscopic pair images such as edge, flat and texture areas and also the blockiness and zero crossing rate within the block of the images are evaluated for artifacts and disparity in this method. The result on our subjective stereoscopic images database indicates that the model performs quite well over a wide rang of image content and distortion levels.

Impact of subjective dataset on the performance of image quality metrics

The interest in objective quality assessment have significantly increased over the past decades. Several objective quality metrics have been proposed and made publicly available, moreover, several subjective quality assessment databases are distributed in order to evaluate and compare the metrics. However, several question arises: are the objective metrics behaviours constant across databases, contents and distortions? how significantly the subjective scores might fluctuate on different displays (i.e. CRT or LCD)? which objective quality metric might best evaluate a given distortion? In this article, we analyse the behaviour of four objective quality metrics (including PSNR) tested on three image databases. We demonstrate that the performances of the quality metrics can strongly fluctuate depending on the database used for testing. We also show the consistency of all metrics for two distinct displays.

Spatial Features Based No Reference Image Quality Assessment for JPEG2000

Perceptual image quality evaluation has become an important issue due to increasing transmission of multimedia contents over the Internet and 3G mobile networks. Most of the no reference perceptual image quality evaluations traditionally attempted to quantify the predefined artifacts of the coded images. Under the assumption that human visual perception is very sensitive to edge information of an image and any kinds of artifacts create pixel distortion, we propose a new philosophy for designing a no reference image quality evaluation model for JPEG2000 images, which uses pixel distortions and edge information. Subjective experiment results on the images are used to train and test the model, which achieve good quality prediction performance.

NR objective continuous video quality assessment model based on frame quality measure

In the current paper, we suggest a new no-reference (NR) objective video quality assessment model. The proposed method can provide continuous quality values at a rate of two scores per second, according to data obtained from subjective assessment tests under the Single Stimulus Continuous Quality Evaluation (SSCQE) method. Moreover, the proposed method is based on a frame quality measure due to the fact that digital video consists of the time series of frame (still) images. Because the frame quality measure alone is not sufficient for an objective video quality assessment, it is necessary to adjust the frame quality value to the information from contents and motion respectively, such as Spatial Infornation (SI) and Temporal Infornation (TI). Our experiments using MPEG-2 coded video sequences indicate that the objective assessment scores obtained by the proposed method are in concordance with the subjective assessment scores.

Objective no-reference stereoscopic image quality prediction based on 2D image features and relative disparity

Stereoscopic images are widely used to enhance the viewing experience of three-dimensional (3D) imaging and communication system. In this paper, we propose an image feature and disparity dependent quality evaluation metric, which incorporates human visible system characteristics. We believe perceived distortions and disparity of any stereoscopic image are strongly dependent on local features, such as edge (i.e., nonplane areas of an image) and nonedge (i.e., plane areas of an image) areas within the image. Therefore, a no-reference perceptual quality assessment method is developed for JPEG coded stereoscopic images based on segmented local features of distortions and disparity. Local feature information such as edge and non-edge area based relative disparity estimation, as well as the blockiness and the edge distortion within the block of images are evaluated in this method. Subjective stereo image database is used for evaluation of the metric. The subjective experiment results indicate that our metric has sufficient prediction performance.

A method to distinguish road surface conditions for car-mounted camera images at night-time

Many conventional distinction methods of road surface condition using car-mounted cameras have been already proposed. However, most of these methods are only effective for daytime and bright conditions. Therefore, we need to expand these methods for impractical night-time that is much more dangerous environment. In this paper, we propose a new distinction method for road surface conditions at night-time, such as dry, wet and snow. This method uses only video information acquired by an inexpensive car-mounted video camera and uses the difference in road surface features for each condition. The image features of the road surface are depending on the illuminant conditions such as street lamps, signal lights, reflections and other lighting sources. Therefore, we analyze the image features based on color information and the presence of other light sources. As a result, the distinction of road surface conditions was achieved with high accuracy, including the areas illuminated by street lamps and other light sources.

Spatio-temporal segmentation based continuous no-reference stereoscopic video quality prediction

In this paper, we propose a no-reference continuous video quality prediction method for MPEG-2 MP@ML coded stereoscopic videos based on spatio-temporal segmentation. Segmented local features such as edge and non-edge areas based spatial artifacts, disparity, and temporal features are measures in this method. Blockiness and blur are considered to measure spatial artifacts for each stereo pair frames. A block based different zero-crossing approach is used for disparity measure. Each temporal segment is evaluated for spatial artifacts and disparity. In this method, temporal features are estimated separately for left and right video sequences based on segmented local features and sub temporal segment. Different weighting factors are then applied for the two different local features to measure the artifacts, disparity, and temporal features of a temporal segment. In order to verify the performance, we conducted subjective experiment on different symmetric and asymmetric coded stereo videos which indicates that our proposed methods prediction performance is quite sufficient.

Objective picture quality scale for video coding

To design the image communication system including human visual system systematically, the objective quality estimation method of the picture based on the model of the human vision is necessary. We develop the objective picture quality scale (PQSvideo) for video coding considering the relation between the physical distortion factors and the psychological picture estimation factors. The obtained cross correlation coefficient between the PQSvideo and the mean opinion score (MOS), becomes 0.978.