Abstract: A method for matching stereo images may calculate a data cost value for each of a plurality of images, calculate a smoothness cost value for each of the plurality of images, and match pixels among the plurality of images based on the data cost value and the smoothness cost value.

Abstract: A method of processing a depth image includes receiving a high-resolution color image and a low-resolution depth image corresponding to the high-resolution color image, generating a feature vector based on a depth distribution of the low-resolution depth image, selecting a filter to upsample the low-resolution depth image by classifying a generated feature vector according to a previously learnt classifier, upsampling the low-resolution depth image by using a selected filter, and outputting an upsampled high-resolution depth image.

Abstract: A method of generating a depth image includes irradiating an object with a light which is generated from a light source, acquiring a plurality of phase difference signals which have different phase differences from one another, by sensing a reflection light reflected from the object, generating a first depth image based on the plurality of phase difference signals, generating a second depth image based on phase difference signals in which a motion artifact has not occurred, among the plurality of phase difference signals, generating a third depth image by combining the first depth image and the second depth image.

Abstract: An apparatus and method for out-focusing a color image based on a depth image, the method including receiving an input of a depth region of interest (ROI) desired to be in focus for performing out-focusing in the depth image, and applying different blur models to pixels corresponding to the depth ROI, and pixels corresponding to a region, other than the depth ROI, in the color image, thereby performing out-focusing on the depth ROI.

Abstract: Provided are a method and an apparatus for image processing. A scene includes a low resolution depth image and a high resolution color image. A depth image up-scaler up-scales the depth image using an interpolation so as to improve the resolution of the depth image. A color-depth similarity based filter removes noise from the up-scaled depth image using the color image and the up-scaled depth image. A boundary sharpening filter converts a boundary between objects in the up-scaled depth image into a clearer boundary. The resolution of the depth image may increase to the level of the resolution of the color image by means of the depth image up-scaler, the color-depth similarity based filter and the boundary sharpening filter, and the depth of each pixel in the depth image may be adjusted.

Abstract: An apparatus and method for processing an image based on a motion of an object, the apparatus including a motion estimator configured to estimate a motion of an object included in a current image, an image determiner configured to determine a neighboring image neighboring the current image based on the motion of the object, and a pixel value determiner configured to determine a pixel value of a hole region neighboring the object based on the neighboring image is provided.

Abstract: An apparatus and method for generating a depth image using transition of light sources may include a surface information determination unit to determine surface information related to a surface of an object, using a plurality of intensity images generated according to position of light sources; and an output depth image generation unit to generate an output depth image, using an original depth image and the surface information.

Abstract: An endoscope using depth information and a method for detecting a polyp based on the endoscope using the depth information are provided. The endoscope using the depth information may generate an irradiated light signal including a visible light, obtain depth information based on the irradiated light signal and a reflected light signal obtained through the irradiated light signal being reflected off of an intestine wall, generate a depth image inside the intestine wall based on the depth information, and detect a polyp located on the intestine wall based on the depth image.

Abstract: Provided is a synthesis system of a time-of-flight (ToF) camera and a stereo camera for reliable wide range depth acquisition and a method therefor. The synthesis system may estimate an error per pixel of a depth image, may calculate a value of a maximum distance multiple per pixel of the depth image using the error per pixel of the depth image, a left color image, and a right color image, and may generate a reconstructed depth image by conducting phase unwrapping on the depth image using the value of the maximum distance multiple per pixel of the depth image.

Abstract: A method and apparatus to generate a volume-panorama image are provided. A method of generating a volume-panorama image includes receiving conversion relationships between volume images, one of the received conversion relationships being between a first volume image of the volume images and a second volume image of the volume images, the second volume image including an area that is common to an area of the first volume image, generating an optimized conversion relationship from the one of the received conversion relationships based on the received conversion relationships, and generating the volume-panorama image based on the generated optimized conversion relationship.

Abstract: Provided is an image processing apparatus and method for detecting a transparent image from an input image. The image processing apparatus may include an image segmenting unit to segment an input image into a plurality of segments, a likelihood determining unit to determine a likelihood that a transparent object is present between adjacent segments among the plurality of segments, and an object detecting unit to detect the transparent object from the input image based on the likelihood.

Abstract: A method for stereo image rectification includes receiving a plurality of images and determining a matrix for performing rectification on the plurality of images based on a first cost function and a second cost function. The first cost function may be associated with a distance of a corresponding pair of points from among the plurality of images. The second cost function may be associated with distortion in a converted image.

Abstract: A method and apparatus for generating a spanning tree, a method and apparatus for stereo matching, a method and apparatus for up-sampling, and a method and apparatus for generating a reference pixel are disclosed, in which a spanning tree may be generated by reference pixels, stereo matching or up-sampling may be performed based on the generated spanning tree, and a reference pixel may be generated based on a stereo video.

Abstract: An image processing apparatus and method for zooming in on a partial area in a three-dimensional (3D) image selects a zoom mode from among a two-dimensional (2D) zoom mode, a 3D zoom mode, and an intermediate dimensional zoom mode between the 2D zoom mode and the 3D zoom mode. The image processing apparatus may include a mode selecting unit to select a zoom mode to be applied to a zoom area in a color image and a depth image among the 2D zoom mode, the 3D zoom mode, and the intermediate dimensional zoom mode between the 2D zoom mode and the 3D zoom mode, and a scaling unit to scale the zoom area using a zoom factor indicating magnification or minification of the zoom area and the selected zoom mode.

Abstract: Provided is an image processing method including: receiving a depth map and a color image with respect to a predetermined scene; projecting a plurality of points included in the depth map, onto the color image; segmenting the color image, onto which the depth map is projected, by using at least two image segmentation methods to generate respective segmentation results, and classifying the plurality of points included in the color image into at least one set according to the respective segmentation results; assigning a depth value to each of the plurality of points of the color image included in the set; and outputting the color image, to which the depth values are assigned.

Abstract: A first calculator of an image processing apparatus may trace a view transform vector between color images of different frames. A second calculator may trace a view transform vector between depth images of different frames. In this example, a third calculator may calculate a view transform vector between the color camera and the depth camera using the view transform vector between the color images and the view transform vector between the depth images.

Abstract: A method for matching stereo images may calculate a data cost value for each of a plurality of images, calculate a smoothness cost value for each of the plurality of images, and match pixels among the plurality of images based on the data cost value and the smoothness cost value.

Abstract: An apparatus for calibrating a multiview image may extract feature points from the multiview image and perform image calibration based on the extracted feature points, track corresponding feature points in temporally successive image frames of a first view image, and perform the image calibration based on pairs of corresponding feature points between the feature points tracked from the first view image and feature points of a second view image.

Abstract: A method and apparatus for processing a depth image that removes noise of a depth image may include a noise estimating unit to estimate noise of a depth image using an amplitude image, a super-pixel generating unit to generate a planar super-pixel based on depth information of the depth image and the noise estimated, and a noise removing unit to remove noise of the depth image using depth information of the depth image and depth information of the super-pixel.

Abstract: Provided is an image processing apparatus and method for detecting a transparent image from an input image. The image processing apparatus may include an image segmenting unit to segment an input image into a plurality of segments, a likelihood determining unit to determine a likelihood that a transparent object is present between adjacent segments among the plurality of segments, and an object detecting unit to detect the transparent object from the input image based on the likelihood.