摘要:
In a method for relative depth of parts of one or more digital images, the digital images are regularized by segmentation, and at least part of the pixels of the images are assigned to respective segments. The realtive motion of the segments for successive images is estimated by image matching. The image features of the segments are regularized by dual segmetation, in which the edges of the segments are found, pixels are assigned to the edges, and dual segments are defined. The relative motion of the dual segments for successive images is estimated by image segment matching in order to determine the relative depth order of the image segments.
摘要:
A video encoder (100) comprises a receiver (101) which receives an uncompressed video signal. An encoding element (103) generates a compressed video signal in accordance with a compression algorithm, such as an MPEG-2 encoding algorithm. In addition, a feature point processor (105) generates feature point data (105) in response to the uncompressed signal, and an output processor (107) generates an output video signal which comprises the compressed video signal and the feature point data. The output signal is received by a receiver (201) of a video signal processor (200). An extraction processor (203) extracts the feature point data and feeds it to a video processor unit (207) which processes the compressed video signal in response to the feature point data. The separate and independent generation of feature point data, such as feature point movement data or tracks using the uncompressed (original) video signal, eliminates or reduces the impact of compression artifacts, inaccuracies and errors on feature point locations and trajectories when these are detected using the compressed signal.
摘要:
There is provided a method of encoding a video signal comprising a sequence of images to generate corresponding encoded video data. The method including the steps of: (a) analyzing the images to identify one or more image segments therein; (b) identifying those of said one or more segments which are substantially not of a spatially stochastic nature and encoding them in a deterministic manner to generate first encoded intermediate data; (c) identifying those of said one or more segments which are of a substantially spatially stochastic nature and encoding them by way of one or more corresponding stochastic model parameters to generate second encoded intermediate data; and (d) merging the first and second intermediate data to generate the encoded video data.
摘要:
The segmentation unit (100,200,300) for assigning a first pixel of a first image of a sequence of images to a segment comprises an assignment unit (102) for assigning a first homogeneity value to the first pixel on basis of the first image; an averaging unit (103) for calculating an average homogeneity value for the first pixel by means of averaging the first homogeneity value and a second homogeneity value being determined for a second pixel of a second image of the sequence of images, with the first pixel and the second pixel related by means of a motion vector, and a comparing unit (106) for comparing the average homogeneity value with a threshold in order to assign the first pixel to the segment.
摘要:
A method of converting of a first set (100a) of initial segments of an image into a second set of updated segments (A′,B′,C′,D′) is disclosed. The method comprises iterative updates of intermediate segments (A,B,C,D) being derived from respective initial segments. Each update comprises determining whether a pixel (300) should be moved from a first intermediate segment (A) to a second intermediate segment (B), on basis of a pixel value of the pixel, on basis of a first parameter of the intermediate segment (A) and on basis of a second parameter of the second intermediate segment (B). The iterative updates are performed on block base. That means that first a number of iterative updates are performed for pixels of a first two-dimensional block of pixels (200) of the image and after that the number of iterative updates are performed for pixels of a second two-dimensional block of pixels (204) of the image.
摘要:
The present invention relates to an apparatus for and a method of propagating depth-related information from a first depth-map (810) associated with a first image (820) to a second depth-map (860) associated with a second image (830), the first and second image being temporally proximate images in an image sequence. The method comprises generating an intermediate depth-map (840) associated with the second image (830) by propagating depth values from the first depth-map (810) to the intermediate depth-map (840) using pixels of the first image (820) and the second image (830), and generating a motion vector (850) using information comprising depth values in a spatial region around a first location in the first depth-map (810) and depth values in a spatial region around a second location in the intermediate depth-map (840) and generating a depth value for the second location in the second depth-map (860) using information comprising the motion vector (850) and the first depth-map (810).
摘要:
A method of rendering a multi-view image comprising a first output image and a second output image on basis of an input image (102) is disclosed. The method comprises: creating a modulation image (100) comprising irregular shaped objects (106-112); modulating pixel values of a portion of the input image (102) on basis of further pixel values of the modulation image (100), resulting into an intermediate image (104); and generating the multi-view image by means of warping the intermediate image on basis of the disparity data.
摘要:
A device for segmenting an image comprises a user input (301, 305, 307) for receiving segment indications and boundary indications for the image. Each segment indication identifies a set of pixel regions and a relationship between the set of pixel regions and an associated segment class of a plurality of segment classes. Each boundary indication identifies a boundary between pixels belonging to different segment classes. A segmentation processor (309) then segments the image into the plurality of segment classes in response to both the number of segment indications and the number of boundary indications. Specifically, a propagation of values linking the pixel regions to segment classes are propagated based on the segment indications with the propagation being constrained (e.g. attenuated or blocked) by the boundary indications. The invention may improve or facilitate interactive image segmentation e.g. for frames of a video signal.
摘要:
The present invention relates to a method and device (580) for encoding three-dimensional video data, the device comprising: a first encoder (505) arranged to encode multiple simultaneous views (501) of a scene from different viewpoints; a second encoder (510) arranged to encode depth information of the scene and a third encoder (515) arranged to encode additional information indicative of a relationship between the multiple views and the depth information and a combiner (520) arranged to combine the encoded information into a representation (513) of the three-dimensional video data. The additional information comprises a group classifier indicating whether or not the depth information and at least one of the multiple views correspond to the same source material for, when corresponding, using the depth information and the at least one of the multiple views to render further views of the scene from further different viewpoints. The invention further related to a method and device (585) for decoding three-dimensional video data, as well as a computer program product comprising program instructions for executing a method according to the present invention.
摘要:
An apparatus (200) comprises a simple detector (201) for detecting a plurality of image points (105, 107, 109, 111) associated with at least one object of the at least one image. The detector does not differentiate between different types of image points. The detector (201) is coupled to a grouping processor (203) which groups the plurality of image points (105, 107, 109, 111) into a group of object points (105, 107), a group of junction points (111) and a group of falsely detected points (109). The apparatus further comprises a processor arrangement 209 for individually processing the image points of the group of object points (105, 107) and the group of junction points (111). The object point process may generate depth information based on dynamic characteristics and the junction point process may generate depth information based on static characteristics. Improved depth information may thus be achieved and a simplified detector may be employed.