Shift-Map Based Stereo Image Retargeting with Disparity Adjustment

Qi, Shaoyu; Ho, Jeffrey

doi:10.1007/978-3-642-37447-0_35

Cited by 8 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The difficulty of keeping disparities within this range during acquisition has spawned a number of computer algorithms capable of automatically adjusting the disparity range of existing stereo content [84,178,128,64,143,116].…”

Section: Excessive Disparitymentioning

confidence: 99%

Causes of discomfort in stereoscopic content: a review

Terzic,

Hansard

2017

Preprint

View full text Add to dashboard Cite

This paper reviews the causes of discomfort in viewing stereoscopic content. These include objective factors, such as misaligned images, as well as subjective factors, such as excessive disparity. Different approaches to the measurement of visual discomfort are also reviewed, in relation to the underlying physiological and psychophysical processes. The importance of understanding these issues, in the context of new display technologies, is emphasized.

show abstract

Section: Excessive Disparitymentioning

confidence: 99%

Causes of discomfort in stereoscopic content: a review

Terzic,

Hansard

2017

Preprint

View full text Add to dashboard Cite

show abstract

“…• Streaming video [3]: the warping method is also used. The saliency map is generated by combining the visual attention map, the line detection, and important objects.• Shift-map stereo image retargeting [10]: in the context of 3D image retargeting, the novel viewpoint advocated is that the geometric consistency in the form of preserving disparity values should not be an overpowering objective formulated as hard constraints. Instead, for maximizing viewing experience and comfort, it is desirable to simultaneously retarget the images as well as adjust the disparity values.…”

mentioning

confidence: 99%

Retargeted Image Quality Assessment: Current Progresses and Future Trends

Deng

Wang

et al. 2014

Visual Signal Quality Assessment

View full text Add to dashboard Cite

Nowadays, the diversity and versatility of the display devices have imposed new demands on digital image processing. The same image needs to be displayed with different resolutions on various devices. The image retargeting approaches [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15] have been proposed to adapt the source images into arbitrary sizes and simultaneously keep the salient content of the source images. These developed retargeting methods, such as warp [2], seam carving [4][5][6], and multi-operator [7], try to preserve the salient shape and content information of the source image, and shrink (or expend) the unimportant regions of the image into the given resolution. With such approaches, the images can be displayed on different screens of different resolutions, which will provide better visual experiences for human viewers. However, for most of these methods, a simple visual comparison was conducted for the results (comparing the results of different retargeting methods based on a small set of images) to demonstrate the effectiveness of the retargeting methods, which is not suitable for online processing. In order to obtain a retargeted image of good quality, quality assessment of retargeted images is needed for maximize the perceptual quality to guide the retargeting process. Therefore, a new challenge of objectively evaluating the retargeted image perceptual quality is issued, where variant resolutions may be presented, the objective shape may be distorted, and some content information may be discarded.As human eyes are the final receivers of the retargeted images, the human subjective opinion is the most reliable value to indicate the image perceptual quality. The subjective opinions are obtained through the subjective testing, where a large number of viewers participate in the subjective test and provide their personal opinions of the image quality on some pre-defined scale. By processing the obtained subjective scores, each image will be assigned a score indicating its perceptual quality. The subjective testing method is time-consuming and expensive, which makes it impractical for most image applications. However, the subjective rating obtained can be recognized as the ground truth of the image perceptual quality. Therefore, the subjective rating scores can be employed to validate the objective quality metrics. Subjective studies can also enable the improvement in the performance of the quality metric towards attaining the ultimate goal of matching human perception. Then the developed quality metric can be utilized to guide the corresponding application. Furthermore, the subjective studies can also benefit the image applications for better perceptual quality experience, specifically improving the perceptual quality of the retargeted image. Consequently there is a great demand of image retargeting database for both subjective rating score acquisition and objective quality metric validation.Objective quality assessment is demanded for not only automatically evaluate the perceptual quality of retar...

show abstract